Showing 1–50 of 159 results for author: So, H

Co-designing a Sub-millisecond Latency Event-based Eye Tracking System with Submanifold Sparse CNN

Authors: Baoheng Zhang, Yizhao Gao, **gyuan Li, Hayden Kwok-Hay So

Abstract: Eye-tracking technology is integral to numerous consumer electronics applications, particularly in the realm of virtual and augmented reality (VR/AR). These applications demand solutions that excel in three crucial aspects: low-latency, low-power consumption, and precision. Yet, achieving optimal performance across all these fronts presents a formidable challenge, necessitating a balance between s… ▽ More Eye-tracking technology is integral to numerous consumer electronics applications, particularly in the realm of virtual and augmented reality (VR/AR). These applications demand solutions that excel in three crucial aspects: low-latency, low-power consumption, and precision. Yet, achieving optimal performance across all these fronts presents a formidable challenge, necessitating a balance between sophisticated algorithms and efficient backend hardware implementations. In this study, we tackle this challenge through a synergistic software/hardware co-design of the system with an event camera. Leveraging the inherent sparsity of event-based input data, we integrate a novel sparse FPGA dataflow accelerator customized for submanifold sparse convolution neural networks (SCNN). The SCNN implemented on the accelerator can efficiently extract the embedding feature vector from each representation of event slices by only processing the non-zero activations. Subsequently, these vectors undergo further processing by a gated recurrent unit (GRU) and a fully connected layer on the host CPU to generate the eye centers. Deployment and evaluation of our system reveal outstanding performance metrics. On the Event-based Eye-Tracking-AIS2024 dataset, our system achieves 81% p5 accuracy, 99.5% p10 accuracy, and 3.71 Mean Euclidean Distance with 0.7 ms latency while only consuming 2.29 mJ per inference. Notably, our solution opens up opportunities for future eye-tracking systems. Code is available at https://github.com/CASR-HKU/ESDA/tree/eye_tracking. △ Less

Submitted 22 April, 2024; originally announced April 2024.

Comments: Accepted to CVPR 2024 workshop, AIS: Vision, Graphics, and AI for Streaming

Event-Based Eye Tracking. AIS 2024 Challenge Survey

Authors: Zuowen Wang, Chang Gao, Zongwei Wu, Marcos V. Conde, Radu Timofte, Shih-Chii Liu, Qinyu Chen, Zheng-jun Zha, Wei Zhai, Han Han, Bohao Liao, Yuliang Wu, Zengyu Wan, Zhong Wang, Yang Cao, Ganchao Tan, **ze Chen, Yan Ru Pei, Sasskia Brüers, Sébastien Crouzet, Douglas McLelland, Oliver Coenen, Baoheng Zhang, Yizhao Gao, **gyuan Li , et al. (14 additional authors not shown)

Abstract: This survey reviews the AIS 2024 Event-Based Eye Tracking (EET) Challenge. The task of the challenge focuses on processing eye movement recorded with event cameras and predicting the pupil center of the eye. The challenge emphasizes efficient eye tracking with event cameras to achieve good task accuracy and efficiency trade-off. During the challenge period, 38 participants registered for the Kaggl… ▽ More This survey reviews the AIS 2024 Event-Based Eye Tracking (EET) Challenge. The task of the challenge focuses on processing eye movement recorded with event cameras and predicting the pupil center of the eye. The challenge emphasizes efficient eye tracking with event cameras to achieve good task accuracy and efficiency trade-off. During the challenge period, 38 participants registered for the Kaggle competition, and 8 teams submitted a challenge factsheet. The novel and diverse methods from the submitted factsheets are reviewed and analyzed in this survey to advance future event-based eye tracking research. △ Less

Submitted 17 April, 2024; originally announced April 2024.

Comments: Qinyu Chen is the corresponding author

One-Bit Target Detection in Collocated MIMO Radar with Colored Background Noise

Authors: Yu-Hang Xiao, David Ramírez, Lei Huang, Xiao Peng Li, Hing Cheung So

Abstract: One-bit sampling has emerged as a promising technique in multiple-input multiple-output (MIMO) radar systems due to its ability to significantly reduce data volume and processing requirements. Nevertheless, current detection methods have not adequately addressed the impact of colored noise, which is frequently encountered in real scenarios. In this paper, we present a novel detection method that a… ▽ More One-bit sampling has emerged as a promising technique in multiple-input multiple-output (MIMO) radar systems due to its ability to significantly reduce data volume and processing requirements. Nevertheless, current detection methods have not adequately addressed the impact of colored noise, which is frequently encountered in real scenarios. In this paper, we present a novel detection method that accounts for colored noise in MIMO radar systems. Specifically, we derive Rao's test by computing the derivative of the likelihood function with respect to the target reflectivity parameter and the Fisher information matrix, resulting in a detector that takes the form of a weighted matched filter. To ensure the constant false alarm rate (CFAR) property, we also consider noise covariance uncertainty and examine its effect on the probability of false alarm. The detection probability is also studied analytically. Simulation results demonstrate that the proposed detector provides considerable performance gains in the presence of colored noise. △ Less

Submitted 26 April, 2024; v1 submitted 11 March, 2024; originally announced March 2024.

A semidefinite programming approach for robust elliptic localization

Authors: Wenxin Xiong, Jiajun He, Zhang-Lei Shi, Keyuan Hu, Hing Cheung So, Chi-Sing Leung

Abstract: This short communication addresses the problem of elliptic localization with outlier measurements, whose occurrences are prevalent in various location-enabled applications and can significantly compromise the positioning performance if not adequately handled. In contrast to the reliance on $M$-estimation adopted in the majority of existing solutions, we take a different path, specifically explorin… ▽ More This short communication addresses the problem of elliptic localization with outlier measurements, whose occurrences are prevalent in various location-enabled applications and can significantly compromise the positioning performance if not adequately handled. In contrast to the reliance on $M$-estimation adopted in the majority of existing solutions, we take a different path, specifically exploring the worst-case robust approximation criterion, to bolster resistance of the elliptic location estimator against outliers. From a geometric standpoint, our method boils down to pinpointing the Chebyshev center of the feasible set determined by the available bistatic ranges with bounded measurement errors. For a practical approach to the associated min-max problem, we convert it into the well-established convex optimization framework of semidefinite programming (SDP). Numerical simulations confirm that our SDP-based technique can outperform a number of existing elliptic localization schemes in terms of positioning accuracy in Gaussian mixture noise, a common type of impulsive interference in the context of range-based localization. △ Less

Submitted 28 January, 2024; originally announced January 2024.

Sparse array design for MIMO radar in multipath scenarios

Authors: Xuchen Li, Ronghao Lin, Hing Cheung So

Abstract: Sparse array designs have focused mostly on angular resolution, peak sidelobe level and directivity factor of virtual arrays for multiple-input multiple-output (MIMO) radar. The notion of the MIMO radar virtual array is based on the direct path assumption in that the direction-of-departure (DOD) and direction-of-arrival (DOA) of the targets are equal. However, the DOD and DOA of targets in multipa… ▽ More Sparse array designs have focused mostly on angular resolution, peak sidelobe level and directivity factor of virtual arrays for multiple-input multiple-output (MIMO) radar. The notion of the MIMO radar virtual array is based on the direct path assumption in that the direction-of-departure (DOD) and direction-of-arrival (DOA) of the targets are equal. However, the DOD and DOA of targets in multipath scenarios are likely to be very different. The identification of multipath targets requires DOD-DOA imaging using the the transmit and receive arrays, not the virtual array. To improve the imaging of both direct path and multipath targets, we introduce several new criteria for MIMO radar sparse linear array (SLA) designs for multipath scenarios. Under the new criteria, we adopt a cyclic optimization strategy under a coordinate descent framework to design the MIMO SLAs. We present several numerical examples to demonstrate the effectiveness of the proposed approaches. △ Less

Submitted 16 January, 2024; originally announced January 2024.

Comments: 5 pages, conference

A Composable Dynamic Sparse Dataflow Architecture for Efficient Event-based Vision Processing on FPGA

Authors: Yizhao Gao, Baoheng Zhang, Yuhao Ding, Hayden Kwok-Hay So

Abstract: Event-based vision represents a paradigm shift in how vision information is captured and processed. By only responding to dynamic intensity changes in the scene, event-based sensing produces far less data than conventional frame-based cameras, promising to springboard a new generation of high-speed, low-power machines for edge intelligence. However, processing such dynamically sparse input origina… ▽ More Event-based vision represents a paradigm shift in how vision information is captured and processed. By only responding to dynamic intensity changes in the scene, event-based sensing produces far less data than conventional frame-based cameras, promising to springboard a new generation of high-speed, low-power machines for edge intelligence. However, processing such dynamically sparse input originated from event cameras efficiently in real time, particularly with complex deep neural networks (DNN), remains a formidable challenge. Existing solutions that employ GPUs and other frame-based DNN accelerators often struggle to efficiently process the dynamically sparse event data, missing the opportunities to improve processing efficiency with sparse data. To address this, we propose ESDA, a composable dynamic sparse dataflow architecture that allows customized DNN accelerators to be constructed rapidly on FPGAs for event-based vision tasks. ESDA is a modular system that is composed of a set of parametrizable modules for each network layer type. These modules share a uniform sparse token-feature interface and can be connected easily to compose an all-on-chip dataflow accelerator on FPGA for each network model. To fully exploit the intrinsic sparsity in event data, ESDA incorporates the use of submanifold sparse convolutions that largely enhance the activation sparsity throughout the layers while simplifying hardware implementation. Finally, a network architecture and hardware implementation co-optimizing framework that allows tradeoffs between accuracy and performance is also presented. Experimental results demonstrate that when compared with existing GPU and hardware-accelerated solutions, ESDA achieves substantial speedup and improvement in energy efficiency across different applications, and it allows much wider design space for real-world deployments. △ Less

Submitted 10 January, 2024; originally announced January 2024.

Comments: Accepted to FPGA'24

Characterising Solar Magnetic Reconnection in Confined and Eruptive Flares

Authors: Kanniah Balamuralikrishna, John Y. H. Soo, Norhaslinda Mohamed Tahrin, Abdul Halim Abdul Aziz

Abstract: Magnetic reconnection is a fundamental mechanism through which energy stored in magnetic fields is released explosively on a massive scale, they could be presented as eruptive or confined flares, depending on their association with coronal mass ejections (CMEs). Several previous works have concluded that there is no correlation between flare duration and flare class, however, their sample sizes ar… ▽ More Magnetic reconnection is a fundamental mechanism through which energy stored in magnetic fields is released explosively on a massive scale, they could be presented as eruptive or confined flares, depending on their association with coronal mass ejections (CMEs). Several previous works have concluded that there is no correlation between flare duration and flare class, however, their sample sizes are skewed towards B and C classes; they hardly represent the higher classes. Therefore, we studied a sample without extreme events in order to determine the correlation between flare duration and flare type (confined and eruptive). We examined $33$ flares with classes between M5 to X5 within $45^{\circ}$ of the disk centres, using data from the Atmospheric Imaging Assembly (AIA) and the Helioseismic and Magnetic Imager (HMI). We find that the linear correlation between flare class against flare duration by full width half maximum (FWHM) in general is weak ($r=0.19$); however, confined flares have a significant correlation ($r=0.58$) compared to eruptive types ($r=0.08$). Also, the confined M class flares' average duration is less than half of the eruptive flares. Similarly, confined flares have a higher correlation ($r=0.89$) than eruptive flares ($r=0.60$) between flare classes against magnetic reconnection flux. In this work, a balanced sample size between flare types is an important strategy for obtaining a reliable quantitative comparison. △ Less

Submitted 20 December, 2023; originally announced December 2023.

Comments: 26 pages, 11 figures

Journal ref: Astrophys. Space Sci. 368 (2023) 94

Machine learning applications in astrophysics: Photometric redshift estimation

Authors: John Y. H. Soo, Ishaq Y. K. Alshuaili, Imdad Mahmud Pathi

Abstract: Machine learning has rose to become an important research tool in the past decade, its application has been expanded to almost if not all disciplines known to mankind. Particularly, the use of machine learning in astrophysics research had a humble beginning in the early 1980s, it has rose and become widely used in many sub-fields today, driven by the vast availability of free astronomical data onl… ▽ More Machine learning has rose to become an important research tool in the past decade, its application has been expanded to almost if not all disciplines known to mankind. Particularly, the use of machine learning in astrophysics research had a humble beginning in the early 1980s, it has rose and become widely used in many sub-fields today, driven by the vast availability of free astronomical data online. In this short review, we narrow our discussion to a single topic in astrophysics - the estimation of photometric redshifts of galaxies and quasars, where we discuss its background, significance, and how machine learning has been used to improve its estimation methods in the past 20 years. We also show examples of some recent machine learning photometric redshift work done in Malaysia, affirming that machine learning is a viable and easy way a develo** nation can contribute towards general research in astronomy and astrophysics. △ Less

Submitted 15 December, 2023; originally announced December 2023.

Comments: 8 pages, 5 figures, published in the proceedings of the First International Conference on Computational Science and Data Analytics (COMDATA), 21-24 November 2021, Kuala Lumpur, Malaysia

Journal ref: AIP Conf. Proc. 2756 (2023) 040001

Random resistive memory-based deep extreme point learning machine for unified visual processing

Authors: Shaocong Wang, Yizhao Gao, Yi Li, Woyu Zhang, Yifei Yu, Bo Wang, Ning Lin, Hegan Chen, Yue Zhang, Yang Jiang, Dingchen Wang, Jia Chen, Peng Dai, Hao Jiang, Peng Lin, Xumeng Zhang, Xiaojuan Qi, Xiaoxin Xu, Hayden So, Zhongrui Wang, Dashan Shang, Qi Liu, Kwang-Ting Cheng, Ming Liu

Abstract: Visual sensors, including 3D LiDAR, neuromorphic DVS sensors, and conventional frame cameras, are increasingly integrated into edge-side intelligent machines. Realizing intensive multi-sensory data analysis directly on edge intelligent machines is crucial for numerous emerging edge applications, such as augmented and virtual reality and unmanned aerial vehicles, which necessitates unified data rep… ▽ More Visual sensors, including 3D LiDAR, neuromorphic DVS sensors, and conventional frame cameras, are increasingly integrated into edge-side intelligent machines. Realizing intensive multi-sensory data analysis directly on edge intelligent machines is crucial for numerous emerging edge applications, such as augmented and virtual reality and unmanned aerial vehicles, which necessitates unified data representation, unprecedented hardware energy efficiency and rapid model training. However, multi-sensory data are intrinsically heterogeneous, causing significant complexity in the system development for edge-side intelligent machines. In addition, the performance of conventional digital hardware is limited by the physically separated processing and memory units, known as the von Neumann bottleneck, and the physical limit of transistor scaling, which contributes to the slowdown of Moore's law. These limitations are further intensified by the tedious training of models with ever-increasing sizes. We propose a novel hardware-software co-design, random resistive memory-based deep extreme point learning machine (DEPLM), that offers efficient unified point set analysis. We show the system's versatility across various data modalities and two different learning tasks. Compared to a conventional digital hardware-based system, our co-design system achieves huge energy efficiency improvements and training cost reduction when compared to conventional systems. Our random resistive memory-based deep extreme point learning machine may pave the way for energy-efficient and training-friendly edge AI across various data modalities and tasks. △ Less

Submitted 14 December, 2023; originally announced December 2023.

Improving Photometric Redshifts by Merging Probability Density Functions from Template-Based and Machine Learning Algorithms

Authors: Ishaq Y. K. Alshuaili, John Y. H. Soo, Mohd Zubir Mat Jafri, Yasmin Rafid

Abstract: This study aims to improve the photometric redshifts (photo-$z$s) of galaxies by integrating two contemporary methods: template-fitting and machine learning. Finding the synergy between these two methods was not a high priority in the past, but now that our computer processing power and observational accuracy have increased, we deem it worth investigating. We compared two methods to improve galaxy… ▽ More This study aims to improve the photometric redshifts (photo-$z$s) of galaxies by integrating two contemporary methods: template-fitting and machine learning. Finding the synergy between these two methods was not a high priority in the past, but now that our computer processing power and observational accuracy have increased, we deem it worth investigating. We compared two methods to improve galaxy photometric redshift estimations by using the algorithms ANNz2 and BPz on different photometric and spectroscopic samples from the Sloan Digital Sky Survey (SDSS). We find that the photometric redshift performance of ANNz2 (machine learning) is better than that of BPz (galactic templates), and with the utilisation of the merging technique we introduced, we see that there is an improvement in photo-$z$ when the two strategies are consolidated, providing improvements in $σ_{RMS}$ and $σ_{68}$ up to [0.0265, 0.0222] in the LRG sample and [0.0471, 0.0471] in the Stripe-82 Sample. This simple demonstration can be used for photo-$z$s of galaxies in fainter and deeper sky surveys, and future work is required to prove its viability in these samples. △ Less

Submitted 28 November, 2023; originally announced November 2023.

Comments: 14 pages, 5 figures

Journal ref: Astron. Lett. 48 (2022) 665-675

The PAU Survey: a new constraint on galaxy formation models using the observed colour redshift relation

Authors: G. Manzoni, C. M. Baugh, P. Norberg, L. Cabayol, J. L. van den Busch, A. Wittje, D. Navarro-Girones, M. Eriksen, P. Fosalba, J. Carretero, F. J. Castander, R. Casas, J. De Vicente, E. Fernandez, J. Garcia-Bellido, E. Gaztanaga, J. C. Helly, H. Hoekstra, H. Hildebrandt, E. J. Gonzalez, S. Koonkor, R. Miquel, C. Padilla, P. Renard, E. Sanchez , et al. (5 additional authors not shown)

Abstract: We use the GALFORM semi-analytical galaxy formation model implemented in the Planck Millennium N-body simulation to build a mock galaxy catalogue on an observer's past lightcone. The mass resolution of this N-body simulation is almost an order of magnitude better than in previous simulations used for this purpose, allowing us to probe fainter galaxies and hence build a more complete mock catalogue… ▽ More We use the GALFORM semi-analytical galaxy formation model implemented in the Planck Millennium N-body simulation to build a mock galaxy catalogue on an observer's past lightcone. The mass resolution of this N-body simulation is almost an order of magnitude better than in previous simulations used for this purpose, allowing us to probe fainter galaxies and hence build a more complete mock catalogue at low redshifts. The high time cadence of the simulation outputs allows us to make improved calculations of galaxy properties and positions in the mock. We test the predictions of the mock against the Physics of the Accelerating Universe Survey, a narrow band imaging survey with highly accurate and precise photometric redshifts, which probes the galaxy population over a lookback time of 8 billion years. We compare the model against the observed number counts, redshift distribution and evolution of the observed colours and find good agreement; these statistics avoid the need for model-dependent processing of the observations. The model produces red and blue populations that have similar median colours to the observations. However, the bimodality of galaxy colours in the model is stronger than in the observations. This bimodality is reduced on including a simple model for errors in the GALFORM photometry. We examine how the model predictions for the observed galaxy colours change when perturbing key model parameters. This exercise shows that the median colours and relative abundance of red and blue galaxies provide constraints on the strength of the feedback driven by supernovae used in the model. △ Less

Submitted 4 March, 2024; v1 submitted 17 November, 2023; originally announced November 2023.

3D Multi-Target Localization Via Intelligent Reflecting Surface: Protocol and Analysis

Authors: Meng Hua, Guangji Chen, Kaitao Meng, Shaodan Ma, Chau Yuen, Hing Cheung So

Abstract: With the emerging environment-aware applications, ubiquitous sensing is expected to play a key role in future networks. In this paper, we study a 3-dimensional (3D) multi-target localization system where multiple intelligent reflecting surfaces (IRSs) are applied to create virtual line-of-sight (LoS) links that bypass the base station (BS) and targets. To fully unveil the fundamental limit of IRS… ▽ More With the emerging environment-aware applications, ubiquitous sensing is expected to play a key role in future networks. In this paper, we study a 3-dimensional (3D) multi-target localization system where multiple intelligent reflecting surfaces (IRSs) are applied to create virtual line-of-sight (LoS) links that bypass the base station (BS) and targets. To fully unveil the fundamental limit of IRS for sensing, we first study a single-target-single-IRS case and propose a novel \textit{two-stage localization protocol} by controlling the on/off state of IRS. To be specific, in the IRS-off stage, we derive the Cramér-Rao bound (CRB) of the azimuth/elevation direction-of-arrival (DoA) of the BS-target link and design a DoA estimator based on the MUSIC algorithm. In the IRS-on stage, the CRB of the azimuth/elevation DoA of the IRS-target link is derived and a simple DoA estimator based on the on-grid IRS beam scanning method is proposed. Particularly, the impact of echo signals reflected by IRS from different paths on sensing performance is analyzed. Moreover, we prove that the single-beam of the IRS is not capable of sensing, but it can be achieved with \textit{multi-beam}. Based on the two obtained DoAs, the 3D single-target location is constructed. We then extend to the multi-target-multi-IRS case and propose an \textit{IRS-adaptive sensing protocol} by controlling the on/off state of multiple IRSs, and a multi-target localization algorithm is developed. Simulation results demonstrate the effectiveness of our scheme and show that sub-meter-level positioning accuracy can be achieved. △ Less

Submitted 28 February, 2024; v1 submitted 24 October, 2023; originally announced October 2023.

Comments: This paper has been submitted to IEEE journal for possible publication

Low-Rank Tensor Completion via Novel Sparsity-Inducing Regularizers

Authors: Zhi-Yong Wang, Hing Cheung So, Abdelhak M. Zoubir

Abstract: To alleviate the bias generated by the l1-norm in the low-rank tensor completion problem, nonconvex surrogates/regularizers have been suggested to replace the tensor nuclear norm, although both can achieve sparsity. However, the thresholding functions of these nonconvex regularizers may not have closed-form expressions and thus iterations are needed, which increases the computational loads. To sol… ▽ More To alleviate the bias generated by the l1-norm in the low-rank tensor completion problem, nonconvex surrogates/regularizers have been suggested to replace the tensor nuclear norm, although both can achieve sparsity. However, the thresholding functions of these nonconvex regularizers may not have closed-form expressions and thus iterations are needed, which increases the computational loads. To solve this issue, we devise a framework to generate sparsity-inducing regularizers with closed-form thresholding functions. These regularizers are applied to low-tubal-rank tensor completion, and efficient algorithms based on the alternating direction method of multipliers are developed. Furthermore, convergence of our methods is analyzed and it is proved that the generated sequences are bounded and any limit point is a stationary point. Experimental results using synthetic and real-world datasets show that the proposed algorithms outperform the state-of-the-art methods in terms of restoration performance. △ Less

Submitted 9 October, 2023; originally announced October 2023.

A framework to generate sparsity-inducing regularizers for enhanced low-rank matrix completion

Authors: Zhi-Yong Wang, Hing Cheung So

Abstract: Applying half-quadratic optimization to loss functions can yield the corresponding regularizers, while these regularizers are usually not sparsity-inducing regularizers (SIRs). To solve this problem, we devise a framework to generate an SIR with closed-form proximity operator. Besides, we specify our framework using several commonly-used loss functions, and produce the corresponding SIRs, which ar… ▽ More Applying half-quadratic optimization to loss functions can yield the corresponding regularizers, while these regularizers are usually not sparsity-inducing regularizers (SIRs). To solve this problem, we devise a framework to generate an SIR with closed-form proximity operator. Besides, we specify our framework using several commonly-used loss functions, and produce the corresponding SIRs, which are then adopted as nonconvex rank surrogates for low-rank matrix completion. Furthermore, algorithms based on the alternating direction method of multipliers are developed. Extensive numerical results show the effectiveness of our methods in terms of recovery performance and runtime. △ Less

Submitted 7 October, 2023; originally announced October 2023.

Robust matrix completion via Novel M-estimator Functions

Authors: Zhi-Yong Wang, Hing Cheung So

Abstract: M-estmators including the Welsch and Cauchy have been widely adopted for robustness against outliers, but they also down-weigh the uncontaminated data. To address this issue, we devise a framework to generate a class of nonconvex functions which only down-weigh outlier-corrupted observations. Our framework is then applied to the Welsch, Cauchy and $\ell_p$-norm functions to produce the correspondi… ▽ More M-estmators including the Welsch and Cauchy have been widely adopted for robustness against outliers, but they also down-weigh the uncontaminated data. To address this issue, we devise a framework to generate a class of nonconvex functions which only down-weigh outlier-corrupted observations. Our framework is then applied to the Welsch, Cauchy and $\ell_p$-norm functions to produce the corresponding robust loss functions. Targeting on the application of robust matrix completion, efficient algorithms based on these functions are developed and their convergence is analyzed. Finally, extensive numerical results demonstrate that the proposed methods are superior to the competitors in terms of recovery accuracy and runtime. △ Less

Submitted 7 October, 2023; originally announced October 2023.

Robust Low-Rank Matrix Completion via a New Sparsity-Inducing Regularizer

Authors: Zhi-Yong Wang, Hing Cheung So, Abdelhak M. Zoubir

Abstract: This paper presents a novel loss function referred to as hybrid ordinary-Welsch (HOW) and a new sparsity-inducing regularizer associated with HOW. We theoretically show that the regularizer is quasiconvex and that the corresponding Moreau envelope is convex. Moreover, the closed-form solution to its Moreau envelope, namely, the proximity operator, is derived. Compared with nonconvex regularizers l… ▽ More This paper presents a novel loss function referred to as hybrid ordinary-Welsch (HOW) and a new sparsity-inducing regularizer associated with HOW. We theoretically show that the regularizer is quasiconvex and that the corresponding Moreau envelope is convex. Moreover, the closed-form solution to its Moreau envelope, namely, the proximity operator, is derived. Compared with nonconvex regularizers like the lp-norm with 0<p<1 that requires iterations to find the corresponding proximity operator, the developed regularizer has a closed-form proximity operator. We apply our regularizer to the robust matrix completion problem, and develop an efficient algorithm based on the alternating direction method of multipliers. The convergence of the suggested method is analyzed and we prove that any generated accumulation point is a stationary point. Finally, experimental results based on synthetic and real-world datasets demonstrate that our algorithm is superior to the state-of-the-art methods in terms of restoration performance. △ Less

Submitted 7 October, 2023; originally announced October 2023.

SpikeMOT: Event-based Multi-Object Tracking with Sparse Motion Features

Authors: Song Wang, Zhu Wang, Can Li, Xiaojuan Qi, Hayden Kwok-Hay So

Abstract: In comparison to conventional RGB cameras, the superior temporal resolution of event cameras allows them to capture rich information between frames, making them prime candidates for object tracking. Yet in practice, despite their theoretical advantages, the body of work on event-based multi-object tracking (MOT) remains in its infancy, especially in real-world settings where events from complex ba… ▽ More In comparison to conventional RGB cameras, the superior temporal resolution of event cameras allows them to capture rich information between frames, making them prime candidates for object tracking. Yet in practice, despite their theoretical advantages, the body of work on event-based multi-object tracking (MOT) remains in its infancy, especially in real-world settings where events from complex background and camera motion can easily obscure the true target motion. In this work, an event-based multi-object tracker, called SpikeMOT, is presented to address these challenges. SpikeMOT leverages spiking neural networks to extract sparse spatiotemporal features from event streams associated with objects. The resulting spike train representations are used to track the object movement at high frequency, while a simultaneous object detector provides updated spatial information of these objects at an equivalent frame rate. To evaluate the effectiveness of SpikeMOT, we introduce DSEC-MOT, the first large-scale event-based MOT benchmark incorporating fine-grained annotations for objects experiencing severe occlusions, frequent trajectory intersections, and long-term re-identification in real-world contexts. Extensive experiments employing DSEC-MOT and another event-based dataset, named FE240hz, demonstrate SpikeMOT's capability to achieve high tracking accuracy amidst challenging real-world scenarios, advancing the state-of-the-art in event-based multi-object tracking. △ Less

Submitted 29 September, 2023; originally announced September 2023.

Network Topology Inference with Sparsity and Laplacian Constraints

Authors: Jiaxi Ying, Xi Han, Rui Zhou, Xiwen Wang, Hing Cheung So

Abstract: We tackle the network topology inference problem by utilizing Laplacian constrained Gaussian graphical models, which recast the task as estimating a precision matrix in the form of a graph Laplacian. Recent research \cite{ying2020nonconvex} has uncovered the limitations of the widely used $\ell_1$-norm in learning sparse graphs under this model: empirically, the number of nonzero entries in the so… ▽ More We tackle the network topology inference problem by utilizing Laplacian constrained Gaussian graphical models, which recast the task as estimating a precision matrix in the form of a graph Laplacian. Recent research \cite{ying2020nonconvex} has uncovered the limitations of the widely used $\ell_1$-norm in learning sparse graphs under this model: empirically, the number of nonzero entries in the solution grows with the regularization parameter of the $\ell_1$-norm; theoretically, a large regularization parameter leads to a fully connected (densest) graph. To overcome these challenges, we propose a graph Laplacian estimation method incorporating the $\ell_0$-norm constraint. An efficient gradient projection algorithm is developed to solve the resulting optimization problem, characterized by sparsity and Laplacian constraints. Through numerical experiments with synthetic and financial time-series datasets, we demonstrate the effectiveness of the proposed method in network topology inference. △ Less

Submitted 2 September, 2023; originally announced September 2023.

Intelligent Reflecting Surface Assisted Localization: Performance Analysis and Algorithm Design

Authors: Meng Hua, Qingqing Wu, Wen Chen, Zesong Fei, Hing Cheung So, Chau Yuen

Abstract: The target sensing/localization performance is fundamentally limited by the line-of-sight link and severe signal attenuation over long distances. This paper considers a challenging scenario where the direct link between the base station (BS) and the target is blocked due to the surrounding blockages and leverages the intelligent reflecting surface (IRS) with some active sensors, termed as \textit{… ▽ More The target sensing/localization performance is fundamentally limited by the line-of-sight link and severe signal attenuation over long distances. This paper considers a challenging scenario where the direct link between the base station (BS) and the target is blocked due to the surrounding blockages and leverages the intelligent reflecting surface (IRS) with some active sensors, termed as \textit{semi-passive IRS}, for localization. To be specific, the active sensors receive echo signals reflected by the target and apply signal processing techniques to estimate the target location. We consider the joint time-of-arrival (ToA) and direction-of-arrival (DoA) estimation for localization and derive the corresponding Cramér-Rao bound (CRB), and then a simple ToA/DoA estimator without iteration is proposed. In particular, the relationships of the CRB for ToA/DoA with the number of frames for IRS beam adjustments, number of IRS reflecting elements, and number of sensors are theoretically analyzed and demystified. Simulation results show that the proposed semi-passive IRS architecture provides sub-meter level positioning accuracy even over a long localization range from the BS to the target and also demonstrate a significant localization accuracy improvement compared to the fully passive IRS architecture. △ Less

Submitted 25 September, 2023; v1 submitted 18 July, 2023; originally announced July 2023.

Comments: The paper has been submitted to IEEE journal for possible publication

SumVg: Total heritability explained by all variants in genome-wide association studies based on summary statistics with standard error estimates

Authors: Hon-Cheong So, Xiao Xue, Pak-Chung Sham

Abstract: Genome-wide association studies (GWAS) are commonly employed to study the genetic basis of complex traits and diseases, and a key question is how much heritability could be explained by all variants in GWAS. One widely used approach that relies on summary statistics only is LD score regression (LDSC), however the approach requires certain assumptions on the SNP effects (all SNPs contribute to heri… ▽ More Genome-wide association studies (GWAS) are commonly employed to study the genetic basis of complex traits and diseases, and a key question is how much heritability could be explained by all variants in GWAS. One widely used approach that relies on summary statistics only is LD score regression (LDSC), however the approach requires certain assumptions on the SNP effects (all SNPs contribute to heritability and each SNP contributes equal variance). More flexible modeling methods may be useful. We previously developed an approach recovering the true z-statistics from a set of observed z-statistics with an empirical Bayes approach, using only summary statistics. However, methods for standard error (SE) estimation are not available yet, limiting the interpretation of results and applicability of the approach. In this study we developed several resampling-based approaches to estimate the SE of SNP-based heritability, including two jackknife and three parametric bootstrap methods. Simulations showed that delete-d-jackknife and parametric bootstrap approaches provide good estimates of the SE. Particularly, the parametric bootstrap approaches yield the lowest root-mean-squared-error (RMSE) of the true SE. In addition, we applied our method to estimate SNP-based heritability of 12 immune-related traits (levels of cytokines and growth factors) to shed light on their genetic architecture. We also implemented the methods to compute the sum of heritability explained and the corresponding SE in an R package SumVg, available at https://github.com/lab-hcso/Estimating-SE-of-total-heritability/ . In conclusion, SumVg may provide a useful alternative tool for SNP heritability and SE estimates, which does not rely on distributional assumptions of SNP effects. △ Less

Submitted 25 June, 2023; originally announced June 2023.

One-Bit Spectrum Sensing for Cognitive Radio

Authors: Pei-Wen Wu, Lei Huang, David Ramírez, Yu-Hang Xiao, Hing Cheung So

Abstract: Spectrum sensing in cognitive radio necessitates effective monitoring of wide bandwidths, which requires high-rate sampling. Traditional spectrum sensing methods employing high-precision analog-to-digital converters (ADCs) result in increased power consumption and expensive hardware costs. In this paper, we explore blind spectrum sensing utilizing one-bit ADCs. We derive a closed-form detector bas… ▽ More Spectrum sensing in cognitive radio necessitates effective monitoring of wide bandwidths, which requires high-rate sampling. Traditional spectrum sensing methods employing high-precision analog-to-digital converters (ADCs) result in increased power consumption and expensive hardware costs. In this paper, we explore blind spectrum sensing utilizing one-bit ADCs. We derive a closed-form detector based on Rao's test and demonstrate its equivalence with the second-order eigenvalue-moment-ratio test. Furthermore, a near-exact distribution based on the moment-based method, and an approximate distribution in the low signal-to-noise ratio (SNR) regime with the use of the central limit theorem, are obtained. Theoretical analysis is then performed and our results show that the performance loss of the proposed detector is approximately $2$ dB ($π/2$) compared to detectors employing $\infty$-bit ADCs when SNR is low. This loss can be compensated for by using approximately $2.47$ ($π^2/4$) times more samples. In addition, we unveil that the efficiency of incoherent accumulation in one-bit detection is the square root of that of coherent accumulation. Simulation results corroborate the correctness of our theoretical calculations. △ Less

Submitted 23 June, 2023; originally announced June 2023.

Robust time-of-arrival localization via ADMM

Authors: Wenxin Xiong, Christian Schindelhauer, Hing Cheung So

Abstract: This article considers the problem of source localization (SL) using possibly unreliable time-of-arrival (TOA) based range measurements. Adopting the strategy of statistical robustification, we formulate TOA SL as minimization of a versatile loss that possesses resistance against the occurrence of outliers. We then present an alternating direction method of multipliers (ADMM) to tackle the nonconv… ▽ More This article considers the problem of source localization (SL) using possibly unreliable time-of-arrival (TOA) based range measurements. Adopting the strategy of statistical robustification, we formulate TOA SL as minimization of a versatile loss that possesses resistance against the occurrence of outliers. We then present an alternating direction method of multipliers (ADMM) to tackle the nonconvex optimization problem in a computationally attractive iterative manner. Moreover, we prove that the solution obtained by the proposed ADMM will correspond to a Karush-Kuhn-Tucker point of the formulation when the algorithm converges, and discuss reasonable assumptions about the robust loss function under which the approach can be theoretically guaranteed to be convergent. Numerical investigations demonstrate the superiority of our method over many existing TOA SL schemes in terms of positioning accuracy and computational simplicity. In particular, the proposed ADMM achieves estimation results with mean square error performance closer to the Cramér-Rao lower bound than its competitors in our simulations of impulsive noise environments. △ Less

Submitted 17 January, 2024; v1 submitted 14 June, 2023; originally announced June 2023.

Comments: Should you have any questions regarding this contribution, please don't hesitate to reach out to me via email at [email protected]

PixelRNN: In-pixel Recurrent Neural Networks for End-to-end-optimized Perception with Neural Sensors

Authors: Haley M. So, Laurie Bose, Piotr Dudek, Gordon Wetzstein

Abstract: Conventional image sensors digitize high-resolution images at fast frame rates, producing a large amount of data that needs to be transmitted off the sensor for further processing. This is challenging for perception systems operating on edge devices, because communication is power inefficient and induces latency. Fueled by innovations in stacked image sensor fabrication, emerging sensor-processors… ▽ More Conventional image sensors digitize high-resolution images at fast frame rates, producing a large amount of data that needs to be transmitted off the sensor for further processing. This is challenging for perception systems operating on edge devices, because communication is power inefficient and induces latency. Fueled by innovations in stacked image sensor fabrication, emerging sensor-processors offer programmability and minimal processing capabilities directly on the sensor. We exploit these capabilities by develo** an efficient recurrent neural network architecture, PixelRNN, that encodes spatio-temporal features on the sensor using purely binary operations. PixelRNN reduces the amount of data to be transmitted off the sensor by a factor of 64x compared to conventional systems while offering competitive accuracy for hand gesture recognition and lip reading tasks. We experimentally validate PixelRNN using a prototype implementation on the SCAMP-5 sensor-processor platform. △ Less

Submitted 11 April, 2023; originally announced April 2023.

One-Bit Covariance Reconstruction with Non-zero Thresholds: Algorithm and Performance Analysis

Authors: Yu-Hang Xiao, Lei Huang, David Ramírez, Cheng Qian, Hing Cheung So

Abstract: Covariance matrix reconstruction is a topic of great significance in the field of one-bit signal processing and has numerous practical applications. Despite its importance, the conventional arcsine law with zero threshold is incapable of recovering the diagonal elements of the covariance matrix. To address this limitation, recent studies have proposed the use of non-zero clip** thresholds. Howev… ▽ More Covariance matrix reconstruction is a topic of great significance in the field of one-bit signal processing and has numerous practical applications. Despite its importance, the conventional arcsine law with zero threshold is incapable of recovering the diagonal elements of the covariance matrix. To address this limitation, recent studies have proposed the use of non-zero clip** thresholds. However, the relationship between the estimation error and the sampling threshold is not yet known. In this paper, we undertake an analysis of the mean squared error by computing the Fisher information matrix for a given threshold. Our results reveal that the optimal threshold can vary considerably, depending on the variances and correlation coefficients. As a result, it is inappropriate to use a constant threshold to encompass parameters that vary widely. To mitigate this issue, we present a recovery scheme that incorporates time-varying thresholds. Our approach differs from existing methods in that it utilizes the exact values of the threshold, rather than its statistical properties, to enhance the estimation performance. Our simulations, including the direction-of-arrival estimation problem, demonstrate the efficacy of the developed scheme, especially in complex scenarios where the covariance elements are widely separated. △ Less

Submitted 29 March, 2023; originally announced March 2023.

DyBit: Dynamic Bit-Precision Numbers for Efficient Quantized Neural Network Inference

Authors: Jiajun Zhou, Jiajun Wu, Yizhao Gao, Yuhao Ding, Chaofan Tao, Boyu Li, Fengbin Tu, Kwang-Ting Cheng, Hayden Kwok-Hay So, Ngai Wong

Abstract: To accelerate the inference of deep neural networks (DNNs), quantization with low-bitwidth numbers is actively researched. A prominent challenge is to quantize the DNN models into low-bitwidth numbers without significant accuracy degradation, especially at very low bitwidths (< 8 bits). This work targets an adaptive data representation with variable-length encoding called DyBit. DyBit can dynamica… ▽ More To accelerate the inference of deep neural networks (DNNs), quantization with low-bitwidth numbers is actively researched. A prominent challenge is to quantize the DNN models into low-bitwidth numbers without significant accuracy degradation, especially at very low bitwidths (< 8 bits). This work targets an adaptive data representation with variable-length encoding called DyBit. DyBit can dynamically adjust the precision and range of separate bit-field to be adapted to the DNN weights/activations distribution. We also propose a hardware-aware quantization framework with a mixed-precision accelerator to trade-off the inference accuracy and speedup. Experimental results demonstrate that the inference accuracy via DyBit is 1.997% higher than the state-of-the-art at 4-bit quantization, and the proposed framework can achieve up to 8.1x speedup compared with the original model. △ Less

Submitted 24 February, 2023; originally announced February 2023.

Globally Optimized TDOA High Frequency Source Localization Based on Quasi-Parabolic Ionosphere Modeling and Collaborative Gradient Projection

Authors: Wenxin Xiong, Christian Schindelhauer, Hing Cheung So

Abstract: We investigate the problem of high frequency (HF) source localization using the time-difference-of-arrival (TDOA) observations of ionosphere-refracted radio rays based on quasi-parabolic (QP) modeling. An unresolved but pertinent issue in such a field is that the existing gradient-type scheme can easily get trapped in local optima for practical use. This will lead to the difficulty in initializing… ▽ More We investigate the problem of high frequency (HF) source localization using the time-difference-of-arrival (TDOA) observations of ionosphere-refracted radio rays based on quasi-parabolic (QP) modeling. An unresolved but pertinent issue in such a field is that the existing gradient-type scheme can easily get trapped in local optima for practical use. This will lead to the difficulty in initializing the algorithm and finally degraded positioning performance if the starting point is inappropriately selected. In this paper, we develop a collaborative gradient projection (GP) algorithm in order to globally solve the highly nonconvex QP-based TDOA HF localization problem. The metaheuristic of particle swarm optimization (PSO) is exploited for information sharing among multiple GP models, each of which is guaranteed to work out a critical point solution to the simplified maximum likelihood formulation. Random mutations are incorporated to avoid the early convergence of PSO. Rather than treating the geolocation of HF transmitter as a pure optimization problem, we further provide workarounds for addressing the possible impairments and challenges when the proposed technique is applied in practice. Numerical results demonstrate the effectiveness of our PSO-assisted re-initialization strategy in achieving the global optimality, and the superiority of our method over its competitor in terms of positioning accuracy. △ Less

Submitted 10 February, 2023; originally announced February 2023.

Comments: This is the accepted version. The final version of this paper has been published in the IEEE Transactions on Aerospace and Electronic Systems. The copyright is with IEEE. This version prevails, as there are unfortunately uncorrected editing mistakes in the final one

Journal ref: in IEEE TAES, vol. 59, no. 1, pp. 580-590, Feb. 2023

Disproving Sum-Difference Co-Array Property

Authors: Jisheng Dai, Hing Cheung So

Abstract: The recently published paper by Gupta and Agrawal [1] exploited the sum-difference co-array (SDCA) to enhance the virtual aperture of sparse arrays. We argue that the key SDCA property established in [1] requires a critical necessary and sufficient condition that is valid for a very rare case only. The recently published paper by Gupta and Agrawal [1] exploited the sum-difference co-array (SDCA) to enhance the virtual aperture of sparse arrays. We argue that the key SDCA property established in [1] requires a critical necessary and sufficient condition that is valid for a very rare case only. △ Less

Submitted 29 January, 2023; originally announced January 2023.

Comments: 2 pages, 1 figure

Unsupervised Mandarin-Cantonese Machine Translation

Authors: Megan Dare, Valentina Fajardo Diaz, Averie Ho Zoen So, Yifan Wang, Shibingfeng Zhang

Abstract: Advancements in unsupervised machine translation have enabled the development of machine translation systems that can translate between languages for which there is not an abundance of parallel data available. We explored unsupervised machine translation between Mandarin Chinese and Cantonese. Despite the vast number of native speakers of Cantonese, there is still no large-scale corpus for the lan… ▽ More Advancements in unsupervised machine translation have enabled the development of machine translation systems that can translate between languages for which there is not an abundance of parallel data available. We explored unsupervised machine translation between Mandarin Chinese and Cantonese. Despite the vast number of native speakers of Cantonese, there is still no large-scale corpus for the language, due to the fact that Cantonese is primarily used for oral communication. The key contributions of our project include: 1. The creation of a new corpus containing approximately 1 million Cantonese sentences, and 2. A large-scale comparison across different model architectures, tokenization schemes, and embedding structures. Our best model trained with character-based tokenization and a Transformer architecture achieved a character-level BLEU of 25.1 when translating from Mandarin to Cantonese and of 24.4 when translating from Cantonese to Mandarin. In this paper we discuss our research process, experiments, and results. △ Less

Submitted 10 January, 2023; originally announced January 2023.

SMILEtrack: SiMIlarity LEarning for Occlusion-Aware Multiple Object Tracking

Authors: Yu-Hsiang Wang, Jun-Wei Hsieh, **-Yang Chen, Ming-Ching Chang, Hung Hin So, Xin Li

Abstract: Despite recent progress in Multiple Object Tracking (MOT), several obstacles such as occlusions, similar objects, and complex scenes remain an open challenge. Meanwhile, a systematic study of the cost-performance tradeoff for the popular tracking-by-detection paradigm is still lacking. This paper introduces SMILEtrack, an innovative object tracker that effectively addresses these challenges by int… ▽ More Despite recent progress in Multiple Object Tracking (MOT), several obstacles such as occlusions, similar objects, and complex scenes remain an open challenge. Meanwhile, a systematic study of the cost-performance tradeoff for the popular tracking-by-detection paradigm is still lacking. This paper introduces SMILEtrack, an innovative object tracker that effectively addresses these challenges by integrating an efficient object detector with a Siamese network-based Similarity Learning Module (SLM). The technical contributions of SMILETrack are twofold. First, we propose an SLM that calculates the appearance similarity between two objects, overcoming the limitations of feature descriptors in Separate Detection and Embedding (SDE) models. The SLM incorporates a Patch Self-Attention (PSA) block inspired by the vision Transformer, which generates reliable features for accurate similarity matching. Second, we develop a Similarity Matching Cascade (SMC) module with a novel GATE function for robust object matching across consecutive video frames, further enhancing MOT performance. Together, these innovations help SMILETrack achieve an improved trade-off between the cost ({\em e.g.}, running speed) and performance (e.g., tracking accuracy) over several existing state-of-the-art benchmarks, including the popular BYTETrack method. SMILETrack outperforms BYTETrack by 0.4-0.8 MOTA and 2.1-2.2 HOTA points on MOT17 and MOT20 datasets. Code is available at https://github.com/**yang1117/SMILEtrack_Official △ Less

Submitted 22 January, 2024; v1 submitted 16 November, 2022; originally announced November 2022.

Comments: Our paper was accepted by AAAI2024

Status and performance of the AMoRE-I experiment on neutrinoless double beta decay

Authors: H. B. Kim, D. H. Ha, E. J. Jeon, J. A. Jeon, H. S. Jo, C. S. Kang, W. G. Kang, H. S. Kim, S. C. Kim, S. G. Kim, S. K. Kim, S. R. Kim, W. T. Kim, Y. D. Kim, Y. H. Kim, D. H. Kwon, E. S. Lee, H. J. Lee, H. S. Lee, J. S. Lee, M. H. Lee, S. W. Lee, Y. C. Lee, D. S. Leonard, H. S. Lim , et al. (10 additional authors not shown)

Abstract: AMoRE is an international project to search for the neutrinoless double beta decay of $^{100}$Mo using a detection technology consisting of magnetic microcalorimeters (MMCs) and molybdenum-based scintillating crystals. Data collection has begun for the current AMORE-I phase of the project, an upgrade from the previous pilot phase. AMoRE-I employs thirteen $^\mathrm{48depl.}$Ca$^{100}$MoO$_4$ cryst… ▽ More AMoRE is an international project to search for the neutrinoless double beta decay of $^{100}$Mo using a detection technology consisting of magnetic microcalorimeters (MMCs) and molybdenum-based scintillating crystals. Data collection has begun for the current AMORE-I phase of the project, an upgrade from the previous pilot phase. AMoRE-I employs thirteen $^\mathrm{48depl.}$Ca$^{100}$MoO$_4$ crystals and five Li$_2$$^{100}$MoO$_4$ crystals for a total crystal mass of 6.2 kg. Each detector module contains a scintillating crystal with two MMC channels for heat and light detection. We report the present status of the experiment and the performance of the detector modules. △ Less

Submitted 5 November, 2022; originally announced November 2022.

Comments: 8 pages, 4 figures, published in Journal of Low Temperature Physics (2022)

Sparse Array Beamformer Design via ADMM

Authors: Hui** Huang, Hing Cheung So, Abdelhak M. Zoubir

Abstract: In this paper, we devise a sparse array design algorithm for adaptive beamforming. Our strategy is based on finding a sparse beamformer weight to maximize the output signal-to-interference-plus-noise ratio (SINR). The proposed method utilizes the alternating direction method of multipliers (ADMM), and admits closed-form solutions at each ADMM iteration. The algorithm convergence properties are ana… ▽ More In this paper, we devise a sparse array design algorithm for adaptive beamforming. Our strategy is based on finding a sparse beamformer weight to maximize the output signal-to-interference-plus-noise ratio (SINR). The proposed method utilizes the alternating direction method of multipliers (ADMM), and admits closed-form solutions at each ADMM iteration. The algorithm convergence properties are analyzed by showing the monotonicity and boundedness of the augmented Lagrangian function. In addition, we prove that the proposed algorithm converges to the set of Karush-Kuhn-Tucker stationary points. Numerical results exhibit its excellent performance, which is comparable to that of the exhaustive search approach, slightly better than those of the state-of-the-art solvers, including the semidefinite relaxation (SDR), its variant (SDR-V), and the successive convex approximation (SCA) approaches, and significantly outperforms several other sparse array design strategies, in terms of output SINR. Moreover, the proposed ADMM algorithm outperforms the SDR, SDR-V, and SCA methods, in terms of computational complexity. △ Less

Submitted 14 October, 2023; v1 submitted 25 August, 2022; originally announced August 2022.

Comments: Updated Appendix D. Accepted by IEEE Transactions on Signal Processing

Convergence Analysis of Consensus-ADMM for General QCQP

Authors: Hui** Huang, Hing Cheung So, Abdelhak M. Zoubir

Abstract: We analyze the convergence properties of the consensus-alternating direction method of multipliers (ADMM) for solving general quadratically constrained quadratic programs. We prove that the augmented Lagrangian function value is monotonically non-increasing as long as the augmented Lagrangian parameter is chosen to be sufficiently large. Simulation results show that the augmented Lagrangian functi… ▽ More We analyze the convergence properties of the consensus-alternating direction method of multipliers (ADMM) for solving general quadratically constrained quadratic programs. We prove that the augmented Lagrangian function value is monotonically non-increasing as long as the augmented Lagrangian parameter is chosen to be sufficiently large. Simulation results show that the augmented Lagrangian function is bounded from below when the matrix in the quadratic term of the objective function is positive definite. In such a case, the consensus-ADMM is convergent. △ Less

Submitted 23 February, 2023; v1 submitted 30 May, 2022; originally announced May 2022.

Comments: 5 figures. This work has been accepted for publication in Signal Processing (Elsevier). Copyright may be transferred without notice, after which this version may no longer be accessible

Automated detection of dark patterns in cookie banners: how to do it poorly and why it is hard to do it any other way

Authors: Than Htut Soe, Cristiana Teixeira Santos, Marija Slavkovik

Abstract: Cookie banners, the pop ups that appear to collect your consent for data collection, are a tempting ground for dark patterns. Dark patterns are design elements that are used to influence the user's choice towards an option that is not in their interest. The use of dark patterns renders consent elicitation meaningless and voids the attempts to improve a fair collection and use of data. Can machine… ▽ More Cookie banners, the pop ups that appear to collect your consent for data collection, are a tempting ground for dark patterns. Dark patterns are design elements that are used to influence the user's choice towards an option that is not in their interest. The use of dark patterns renders consent elicitation meaningless and voids the attempts to improve a fair collection and use of data. Can machine learning be used to automatically detect the presence of dark patterns in cookie banners? In this work, a dataset of cookie banners of 300 news websites was used to train a prediction model that does exactly that. The machine learning pipeline we used includes feature engineering, parameter search, training a Gradient Boosted Tree classifier and evaluation. The accuracy of the trained model is promising, but allows a lot of room for improvement. We provide an in-depth analysis of the interdisciplinary challenges that automated dark pattern detection poses to artificial intelligence. The dataset and all the code created using machine learning is available at the url to repository removed for review. △ Less

Submitted 21 April, 2022; originally announced April 2022.

On Generalisation of Isotropic Central Difference for Higher Order Approximation of Fractional Laplacian

Authors: Pui Ho Lam, Hing Cheung So

Abstract: The study of generalising the central difference for integer order Laplacian to fractional order is discussed in this paper. Analysis shows that, in contrary to the conclusion of a previous study, difference stencils evaluated through fast Fourier transform prevents the convergence of the solution of fractional Laplacian. We propose a composite quadrature rule in order to efficiently evaluate the… ▽ More The study of generalising the central difference for integer order Laplacian to fractional order is discussed in this paper. Analysis shows that, in contrary to the conclusion of a previous study, difference stencils evaluated through fast Fourier transform prevents the convergence of the solution of fractional Laplacian. We propose a composite quadrature rule in order to efficiently evaluate the stencil coefficients with the required convergence rate in order to guarantee convergence of the solution. Furthermore, we propose the use of generalised higher order lattice Boltzmann method to generate stencils which can approximate fractional Laplacian with higher order convergence speed and error isotropy. We also review the formulation of the lattice Boltzmann method and discuss the explicit sparse solution formulated using Smolyak's algorithm, as well as the method for the evaluation of the Hermite polynomials for efficient generation of the higher order stencils. Numerical experiments are carried out to verify the error analysis and formulations. △ Less

Submitted 13 February, 2022; originally announced February 2022.

Low-Rank and Row-Sparse Decomposition for Joint DOA Estimation and Distorted Sensor Detection

Authors: Hui** Huang, Qi Liu, Hing Cheung So, Abdelhak M. Zoubir

Abstract: Distorted sensors could occur randomly and may lead to the breakdown of a sensor array system. We consider an array model within which a small number of sensors are distorted by unknown sensor gain and phase errors. With such an array model, the problem of joint direction-of-arrival (DOA) estimation and distorted sensor detection is investigated and the problem is formulated under the framework of… ▽ More Distorted sensors could occur randomly and may lead to the breakdown of a sensor array system. We consider an array model within which a small number of sensors are distorted by unknown sensor gain and phase errors. With such an array model, the problem of joint direction-of-arrival (DOA) estimation and distorted sensor detection is investigated and the problem is formulated under the framework of low-rank and row-sparse decomposition. We derive an iteratively reweighted least squares (IRLS) algorithm to solve the resulting problem in both noiseless and noisy cases. The convergence property of the IRLS algorithm is analyzed by means of the monotonicity and boundedness of the objective function. Extensive simulations are conducted regarding parameter selection, convergence speed, computational complexity, and performances of DOA estimation as well as distorted sensor detection. Even though the IRLS algorithm is slightly worse than the alternating direction method of multipliers in detecting the distorted sensors, the results show that our approach outperforms several state-of-the-art techniques in terms of convergence speed, computational cost, and DOA estimation performance. △ Less

Submitted 25 August, 2022; v1 submitted 2 February, 2022; originally announced February 2022.

Comments: This work has been submitted to the IEEE for possible publication. Copyright may be transferred without notice, after which this version may no longer be accessible

Off-Grid Direction-of-Arrival Estimation Using Second-Order Taylor Approximation

Authors: Hui** Huang, Hing Cheung So, Abdelhak M. Zoubir

Abstract: The problem of off-grid direction-of-arrival (DOA) estimation is investigated. We develop a grid-based method to jointly estimate the closest spatial frequency (the sine of DOA) grids, and the gaps between the estimated grids and the corresponding frequencies. By using a second-order Taylor approximation, the data model under the framework of joint-sparse representation is formulated. We point out… ▽ More The problem of off-grid direction-of-arrival (DOA) estimation is investigated. We develop a grid-based method to jointly estimate the closest spatial frequency (the sine of DOA) grids, and the gaps between the estimated grids and the corresponding frequencies. By using a second-order Taylor approximation, the data model under the framework of joint-sparse representation is formulated. We point out an important property of the signals of interest in the model, namely the proportionality relationship, which is empirically demonstrated to be useful in the sense that it increases the probability of the mixing matrix satisfying the block restricted isometry property. Simulation examples demonstrate the effectiveness and superiority of the proposed method against several state-of-the-art grid-based approaches. △ Less

Submitted 26 February, 2022; v1 submitted 10 December, 2021; originally announced December 2021.

Comments: 20 pages, 9 figures, accepted for publication in Signal Processing (Elsevier)

Journal ref: journal = {Signal Processing}, pages = {108513}, year = {2022}, issn = {0165-1684},

MantissaCam: Learning Snapshot High-dynamic-range Imaging with Perceptually-based In-pixel Irradiance Encoding

Authors: Haley M. So, Julien N. P. Martel, Piotr Dudek, Gordon Wetzstein

Abstract: The ability to image high-dynamic-range (HDR) scenes is crucial in many computer vision applications. The dynamic range of conventional sensors, however, is fundamentally limited by their well capacity, resulting in saturation of bright scene parts. To overcome this limitation, emerging sensors offer in-pixel processing capabilities to encode the incident irradiance. Among the most promising encod… ▽ More The ability to image high-dynamic-range (HDR) scenes is crucial in many computer vision applications. The dynamic range of conventional sensors, however, is fundamentally limited by their well capacity, resulting in saturation of bright scene parts. To overcome this limitation, emerging sensors offer in-pixel processing capabilities to encode the incident irradiance. Among the most promising encoding schemes is modulo wrap**, which results in a computational photography problem where the HDR scene is computed by an irradiance unwrap** algorithm from the wrapped low-dynamic-range (LDR) sensor image. Here, we design a neural network--based algorithm that outperforms previous irradiance unwrap** methods and we design a perceptually inspired "mantissa" encoding scheme that more efficiently wraps an HDR scene into an LDR sensor. Combined with our reconstruction framework, MantissaCam achieves state-of-the-art results among modulo-type snapshot HDR imaging approaches. We demonstrate the efficacy of our method in simulation and show benefits of our algorithm on modulo images captured with a prototype implemented with a programmable sensor. △ Less

Submitted 20 April, 2022; v1 submitted 9 December, 2021; originally announced December 2021.

Nonlinear Tensor Ring Network

Authors: Xiao Peng Li, Qi Liu, Hing Cheung So

Abstract: The state-of-the-art deep neural networks (DNNs) have been widely applied for various real-world applications, and achieved significant performance for cognitive problems. However, the increment of DNNs' width and depth in architecture results in a huge amount of parameters to challenge the storage and memory cost, limiting to the usage of DNNs on resource-constrained platforms, such as portable d… ▽ More The state-of-the-art deep neural networks (DNNs) have been widely applied for various real-world applications, and achieved significant performance for cognitive problems. However, the increment of DNNs' width and depth in architecture results in a huge amount of parameters to challenge the storage and memory cost, limiting to the usage of DNNs on resource-constrained platforms, such as portable devices. By converting redundant models into compact ones, compression technique appears to be a practical solution to reducing the storage and memory consumption. In this paper, we develop a nonlinear tensor ring network (NTRN) in which both fullyconnected and convolutional layers are compressed via tensor ring decomposition. Furthermore, to mitigate the accuracy loss caused by compression, a nonlinear activation function is embedded into the tensor contraction and convolution operations inside the compressed layer. Experimental results demonstrate the effectiveness and superiority of the proposed NTRN for image classification using two basic neural networks, LeNet-5 and VGG-11 on three datasets, viz. MNIST, Fashion MNIST and Cifar-10. △ Less

Submitted 11 November, 2021; originally announced November 2021.

AI video editing tools. What editors want and how far is AI from delivering?

Authors: Than Htut Soe

Abstract: Video editing can be a very tedious task, so unsurprisingly Artificial Intelligence has been increasingly used to streamline the workflow or automate away tedious tasks. However, it is very difficult to get an overview of what intelligent video editing tools are in the research literature and needs for automation from the video editors. So, we identified the field of intelligent video editing tool… ▽ More Video editing can be a very tedious task, so unsurprisingly Artificial Intelligence has been increasingly used to streamline the workflow or automate away tedious tasks. However, it is very difficult to get an overview of what intelligent video editing tools are in the research literature and needs for automation from the video editors. So, we identified the field of intelligent video editing tools in research, and we survey the opinions of professional video editors. We have also summarized current state of the art in artificial intelligence research with the intention of identifying what are the possibilities and current technical limits towards truly intelligent video editing tools. The findings contribute towards understanding of the field of intelligent video editing tools, highlights unaddressed automation needs by the survey and provides general suggestions for further research in intelligent video editing tools. △ Less

Submitted 16 September, 2021; originally announced September 2021.

ACM Class: H.5; I.2

Arbitrary order of convergence for Riesz fractional derivative via central difference method

Authors: Pui Ho Lam, Hing Cheung So, Cheung Fat Chan

Abstract: We propose a novel method to compute a finite difference stencil for Riesz derivative for artibitrary speed of convergence. This method is based on applying a pre-filter to the Grünwald-Letnikov type central difference stencil. The filter is obtained by solving for the inverse of a symmetric Vandemonde matrix and exploiting the relationship between the Taylor's series coefficients and fast Fourier… ▽ More We propose a novel method to compute a finite difference stencil for Riesz derivative for artibitrary speed of convergence. This method is based on applying a pre-filter to the Grünwald-Letnikov type central difference stencil. The filter is obtained by solving for the inverse of a symmetric Vandemonde matrix and exploiting the relationship between the Taylor's series coefficients and fast Fourier transform. The filter costs O\left(N^{2}\right) operations to evaluate for O\left(h^{N}\right) of convergence, where h is the sampling distance. The higher convergence speed should more than offset the overhead with the requirement of the number of nodal points for a desired error tolerance significantly reduced. The benefit of progressive generation of the stencil coefficients for adaptive grid size for dynamic problems with the Grünwald-Letnikov type difference scheme is also kept because of the application of filtering. The higher convergence rate is verified through numerical experiments. △ Less

Submitted 8 August, 2021; originally announced August 2021.

Comments: 14 pages, 2 figures

MSC Class: 65B99; 65M06; 65R10; 26A33

Characterising Improvements in Photometric Redshift Probability Density Functions with Galaxy Morphology

Authors: John Y. H. Soo, Benjamin Joachimi

Abstract: In this work, we studied the impact of galaxy morphology on photometric redshift (photo-$z$) probability density functions (PDFs). By including galaxy morphological parameters like the radius, axis-ratio, surface brightness and the Sérsic index in addition to the $ugriz$ broadbands as input parameters, we used the machine learning photo-$z$ algorithm ANNz2 to train and test on galaxies from the Ca… ▽ More In this work, we studied the impact of galaxy morphology on photometric redshift (photo-$z$) probability density functions (PDFs). By including galaxy morphological parameters like the radius, axis-ratio, surface brightness and the Sérsic index in addition to the $ugriz$ broadbands as input parameters, we used the machine learning photo-$z$ algorithm ANNz2 to train and test on galaxies from the Canada-France-Hawaii Telescope Stripe-82 (CS82) Survey. Metrics like the continuous ranked probability score (CRPS), probability integral transform (PIT), Bayesian odds parameter, and even the width and height of the PDFs were evaluated, and the results were compared when different number of input parameters were used during the training process. We find improvements in the CRPS and width of the PDFs when galaxy morphology has been added to the training, and the improvement is larger especially when the number of broadband magnitudes are lacking. △ Less

Submitted 7 July, 2021; originally announced July 2021.

Comments: 4 pages, 1 figure, published in the Proceedings of the 14th Asia-Pacific Physics Conference (APPC) held in Kuching, Malaysia on 17-22 November 2019

Journal ref: AIP Conference Proceedings 2319 (2021), 040002

Exploiting Elasticity in Tensor Ranks for Compressing Neural Networks

Authors: Jie Ran, Rui Lin, Hayden K. H. So, Graziano Chesi, Ngai Wong

Abstract: Elasticities in depth, width, kernel size and resolution have been explored in compressing deep neural networks (DNNs). Recognizing that the kernels in a convolutional neural network (CNN) are 4-way tensors, we further exploit a new elasticity dimension along the input-output channels. Specifically, a novel nuclear-norm rank minimization factorization (NRMF) approach is proposed to dynamically and… ▽ More Elasticities in depth, width, kernel size and resolution have been explored in compressing deep neural networks (DNNs). Recognizing that the kernels in a convolutional neural network (CNN) are 4-way tensors, we further exploit a new elasticity dimension along the input-output channels. Specifically, a novel nuclear-norm rank minimization factorization (NRMF) approach is proposed to dynamically and globally search for the reduced tensor ranks during training. Correlation between tensor ranks across multiple layers is revealed, and a graceful tradeoff between model size and accuracy is obtained. Experiments then show the superiority of NRMF over the previous non-elastic variational Bayesian matrix factorization (VBMF) scheme. △ Less

Submitted 10 May, 2021; originally announced May 2021.

Comments: 8 pages, 5 figures

HAO: Hardware-aware neural Architecture Optimization for Efficient Inference

Authors: Zhen Dong, Yizhao Gao, Qi**g Huang, John Wawrzynek, Hayden K. H. So, Kurt Keutzer

Abstract: Automatic algorithm-hardware co-design for DNN has shown great success in improving the performance of DNNs on FPGAs. However, this process remains challenging due to the intractable search space of neural network architectures and hardware accelerator implementation. Differing from existing hardware-aware neural architecture search (NAS) algorithms that rely solely on the expensive learning-based… ▽ More Automatic algorithm-hardware co-design for DNN has shown great success in improving the performance of DNNs on FPGAs. However, this process remains challenging due to the intractable search space of neural network architectures and hardware accelerator implementation. Differing from existing hardware-aware neural architecture search (NAS) algorithms that rely solely on the expensive learning-based approaches, our work incorporates integer programming into the search algorithm to prune the design space. Given a set of hardware resource constraints, our integer programming formulation directly outputs the optimal accelerator configuration for map** a DNN subgraph that minimizes latency. We use an accuracy predictor for different DNN subgraphs with different quantization schemes and generate accuracy-latency pareto frontiers. With low computational cost, our algorithm can generate quantized networks that achieve state-of-the-art accuracy and hardware performance on Xilinx Zynq (ZU3EG) FPGA for image classification on ImageNet dataset. The solution searched by our algorithm achieves 72.5% top-1 accuracy on ImageNet at framerate 50, which is 60% faster than MnasNet and 135% faster than FBNet with comparable accuracy. △ Less

Submitted 26 April, 2021; originally announced April 2021.

Journal ref: FCCM 2021

The PAU Survey: narrowband photometric redshifts using Gaussian processes

Authors: John Y. H. Soo, Benjamin Joachimi, Martin Eriksen, Małgorzata Siudek, Alex Alarcon, Laura Cabayol, Jorge Carretero, Ricard Casas, Francisco J. Castander, Enrique Fernández, Juan Garciá-Bellido, Enrique Gaztanaga, Hendrik Hildebrandt, Henk Hoekstra, Ramon Miquel, Cristobal Padilla, Eusebio Sánchez, Santiago Serrano, Pau Tallada-Crespí

Abstract: We study the performance of the hybrid template-machine-learning photometric redshift (photo-$z$) algorithm Delight, which uses Gaussian processes, on a subset of the early data release of the Physics of the Accelerating Universe Survey (PAUS). We calibrate the fluxes of the $40$ PAUS narrow bands with $6$ broadband fluxes ($uBVriz$) in the COSMOS field using three different methods, including a n… ▽ More We study the performance of the hybrid template-machine-learning photometric redshift (photo-$z$) algorithm Delight, which uses Gaussian processes, on a subset of the early data release of the Physics of the Accelerating Universe Survey (PAUS). We calibrate the fluxes of the $40$ PAUS narrow bands with $6$ broadband fluxes ($uBVriz$) in the COSMOS field using three different methods, including a new method which utilises the correlation between the apparent size and overall flux of the galaxy. We use a rich set of empirically derived galaxy spectral templates as guides to train the Gaussian process, and we show that our results are competitive with other standard photometric redshift algorithms. Delight achieves a photo-$z$ $68$th percentile error of $σ_{68}=0.0081(1+z)$ without any quality cut for galaxies with $i_\mathrm{auto}<22.5$ as compared to $0.0089(1+z)$ and $0.0202(1+z)$ for the BPz and ANNz2 codes, respectively. Delight is also shown to produce more accurate probability distribution functions for individual redshift estimates than BPz and ANNz2. Common photo-$z$ outliers of Delight and BCNz2 (previously applied to PAUS) are found to be primarily caused by outliers in the narrowband fluxes, with a small number of cases potentially indicating spectroscopic redshift failures in the reference sample. In the process, we introduce performance metrics derived from the results of BCNz2 and Delight, allowing us to achieve a photo-$z$ quality of $σ_{68}<0.0035(1+z)$ at a magnitude of $i_\mathrm{auto}<22.5$ while kee** $50$ per cent objects of the galaxy sample. △ Less

Submitted 23 March, 2021; v1 submitted 11 January, 2021; originally announced January 2021.

Comments: 19 pages, 16 figures, accepted by MNRAS

Mix and Match: A Novel FPGA-Centric Deep Neural Network Quantization Framework

Authors: Sung-En Chang, Yanyu Li, Mengshu Sun, Runbin Shi, Hayden K. -H. So, Xuehai Qian, Yanzhi Wang, Xue Lin

Abstract: Deep Neural Networks (DNNs) have achieved extraordinary performance in various application domains. To support diverse DNN models, efficient implementations of DNN inference on edge-computing platforms, e.g., ASICs, FPGAs, and embedded systems, are extensively investigated. Due to the huge model size and computation amount, model compression is a critical step to deploy DNN models on edge devices.… ▽ More Deep Neural Networks (DNNs) have achieved extraordinary performance in various application domains. To support diverse DNN models, efficient implementations of DNN inference on edge-computing platforms, e.g., ASICs, FPGAs, and embedded systems, are extensively investigated. Due to the huge model size and computation amount, model compression is a critical step to deploy DNN models on edge devices. This paper focuses on weight quantization, a hardware-friendly model compression approach that is complementary to weight pruning. Unlike existing methods that use the same quantization scheme for all weights, we propose the first solution that applies different quantization schemes for different rows of the weight matrix. It is motivated by (1) the distribution of the weights in the different rows are not the same; and (2) the potential of achieving better utilization of heterogeneous FPGA hardware resources. To achieve that, we first propose a hardware-friendly quantization scheme named sum-of-power-of-2 (SP2) suitable for Gaussian-like weight distribution, in which the multiplication arithmetic can be replaced with logic shifter and adder, thereby enabling highly efficient implementations with the FPGA LUT resources. In contrast, the existing fixed-point quantization is suitable for Uniform-like weight distribution and can be implemented efficiently by DSP. Then to fully explore the resources, we propose an FPGA-centric mixed scheme quantization (MSQ) with an ensemble of the proposed SP2 and the fixed-point schemes. Combining the two schemes can maintain, or even increase accuracy due to better matching with weight distributions. △ Less

Submitted 11 December, 2020; v1 submitted 8 December, 2020; originally announced December 2020.

Comments: Accepted by High-Performance Computer Architecture (HPCA'2021)

MSC Class: 68T07

Modelling Complex Survey Data Using R, SAS, SPSS and Stata: A Comparison Using CLSA Datasets

Authors: Hon Yiu So, Urun Erbas Oz, Lauren Griffith, Susan Kirkland, **hua Ma, Parminder Raina, Nazmul Sohel, Mary E. Thompson, Christina Wolfson, Changbao Wu

Abstract: The R software has become popular among researchers due to its flexibility and open-source nature. However, researchers in the fields of public health and epidemiological studies are more customary to commercial statistical softwares such as SAS, SPSS and Stata. This paper provides a comprehensive comparison on analysis of health survey data using the R survey package, SAS, SPSS and Stata. We desc… ▽ More The R software has become popular among researchers due to its flexibility and open-source nature. However, researchers in the fields of public health and epidemiological studies are more customary to commercial statistical softwares such as SAS, SPSS and Stata. This paper provides a comprehensive comparison on analysis of health survey data using the R survey package, SAS, SPSS and Stata. We describe detailed R codes and procedures for other software packages on commonly encountered statistical analyses, such as estimation of population means and regression analysis, using datasets from the Canadian Longitudinal Study on Aging (CLSA). It is hoped that the paper stimulates interest among health science researchers to carry data analysis using R and also serves as a cookbook for statistical analysis using different software packages. △ Less

Submitted 24 October, 2020; v1 submitted 19 October, 2020; originally announced October 2020.

Comments: There is a data usage issue with the paper and it is requested by the data owner (CLSA) to withdraw the paper at this time

MSC Class: 62-04 (Primary); 62D05 (secondary)

NITI: Training Integer Neural Networks Using Integer-only Arithmetic

Authors: Maolin Wang, Seyedramin Rasoulinezhad, Philip H. W. Leong, Hayden K. H. So

Abstract: While integer arithmetic has been widely adopted for improved performance in deep quantized neural network inference, training remains a task primarily executed using floating point arithmetic. This is because both high dynamic range and numerical accuracy are central to the success of most modern training algorithms. However, due to its potential for computational, storage and energy advantages i… ▽ More While integer arithmetic has been widely adopted for improved performance in deep quantized neural network inference, training remains a task primarily executed using floating point arithmetic. This is because both high dynamic range and numerical accuracy are central to the success of most modern training algorithms. However, due to its potential for computational, storage and energy advantages in hardware accelerators, neural network training methods that can be implemented with low precision integer-only arithmetic remains an active research challenge. In this paper, we present NITI, an efficient deep neural network training framework that stores all parameters and intermediate values as integers, and computes exclusively with integer arithmetic. A pseudo stochastic rounding scheme that eliminates the need for external random number generation is proposed to facilitate conversion from wider intermediate results to low precision storage. Furthermore, a cross-entropy loss backpropagation scheme computed with integer-only arithmetic is proposed. A proof-of-concept open-source software implementation of NITI that utilizes native 8-bit integer operations in modern GPUs to achieve end-to-end training is presented. When compared with an equivalent training setup implemented with floating point storage and arithmetic, NITI achieves negligible accuracy degradation on the MNIST and CIFAR10 datasets using 8-bit integer storage and computation. On ImageNet, 16-bit integers are needed for weight accumulation with an 8-bit datapath. This achieves training results comparable to all-floating-point implementations. △ Less

Submitted 11 February, 2022; v1 submitted 28 September, 2020; originally announced September 2020.

Neurodynamic TDOA localization with NLOS mitigation via maximum correntropy criterion

Authors: Wenxin Xiong, Christian Schindelhauer, Hing Cheung So, Junli Liang, Zhi Wang

Abstract: In this paper, we exploit the maximum correntropy criterion (MCC) to robustify the traditional time-difference-of-arrival (TDOA) location estimator in the presence of non-line-of-sight (NLOS) propagation conditions. For the sake of statistical efficiency, the correntropy-based robust loss is imposed on the underlying time-of-arrival composition via joint estimation of the source position and onset… ▽ More In this paper, we exploit the maximum correntropy criterion (MCC) to robustify the traditional time-difference-of-arrival (TDOA) location estimator in the presence of non-line-of-sight (NLOS) propagation conditions. For the sake of statistical efficiency, the correntropy-based robust loss is imposed on the underlying time-of-arrival composition via joint estimation of the source position and onset time, instead of the TDOA counterpart generated in the postprocessing of sensor-collected timestamps. We then employ a neurodynamic optimization approach to tackle the highly nonconvex MCC formulation. Furthermore, we examine the local stability of equilibrium for the corresponding projection-type neural network model. Simulation investigations in representative NLOS propagation scenarios demonstrate that our neurodynamic robust TDOA localization solution is capable of outperforming several existing schemes in terms of positioning accuracy. △ Less

Submitted 9 November, 2021; v1 submitted 14 September, 2020; originally announced September 2020.

Comments: Submitted to DSP

Maximum correntropy criterion for robust TOA-based localization in NLOS environments

Authors: Wenxin Xiong, Christian Schindelhauer, Hing Cheung So, Zhi Wang

Abstract: We investigate the problem of time-of-arrival (TOA) based localization under possible non-line-of-sight (NLOS) propagation conditions. To robustify the squared-range-based location estimator, we follow the maximum correntropy criterion, essentially the Welsch $M$-estimator with a redescending influence function which behaves like $\ell_0$-minimization towards the grossly biased measurements, to de… ▽ More We investigate the problem of time-of-arrival (TOA) based localization under possible non-line-of-sight (NLOS) propagation conditions. To robustify the squared-range-based location estimator, we follow the maximum correntropy criterion, essentially the Welsch $M$-estimator with a redescending influence function which behaves like $\ell_0$-minimization towards the grossly biased measurements, to derive the formulation. The half-quadratic technique is then applied to settle the resulting optimization problem in an alternating maximization (AM) manner. By construction, the major computational challenge at each AM iteration boils down to handling an easily solvable generalized trust region subproblem. It is worth noting that the implementation of our localization method requires nothing but merely the TOA-based range measurements and sensor positions as prior information. Simulation and experimental results demonstrate the competence of the presented scheme in outperforming several state-of-the-art approaches in terms of positioning accuracy, especially in scenarios where the percentage of NLOS paths is not large enough. △ Less

Submitted 10 September, 2021; v1 submitted 13 September, 2020; originally announced September 2020.

Comments: Published in CSSP

Inverse NN Modelling of a Piezoelectric Stage with Dominant Variable

Authors: Gangfeng Yan, Hang Jian Soo, Khalid Abidi, Jian-Xin Xu

Abstract: This paper presents an approach for develo** a neural network inverse model of a piezoelectric positioning stage, which exhibits rate-dependent, asymmetric hysteresis. It is shown that using both the velocity and the acceleration as inputs results in over-fitting. To overcome this, a rough analytical model of the actuator is derived and by measuring its response to excitation, the velocity signa… ▽ More This paper presents an approach for develo** a neural network inverse model of a piezoelectric positioning stage, which exhibits rate-dependent, asymmetric hysteresis. It is shown that using both the velocity and the acceleration as inputs results in over-fitting. To overcome this, a rough analytical model of the actuator is derived and by measuring its response to excitation, the velocity signal is identified as the dominant variable. By setting the input space of the neural network to only the dominant variable, an inverse model with good predictive ability is obtained. Training of the network is accomplished using the Levenberg-Marquardt algorithm. Finally, the effectiveness of the proposed approach is experimentally demonstrated. △ Less

Submitted 30 July, 2020; originally announced July 2020.