-
DenRAM: Neuromorphic Dendritic Architecture with RRAM for Efficient Temporal Processing with Delays
Authors:
Simone DAgostino,
Filippo Moro,
Tristan Torchet,
Yigit Demirag,
Laurent Grenouillet,
Giacomo Indiveri,
Elisa Vianello,
Melika Payvand
Abstract:
An increasing number of neuroscience studies are highlighting the importance of spatial dendritic branching in pyramidal neurons in the brain for supporting non-linear computation through localized synaptic integration. In particular, dendritic branches play a key role in temporal signal processing and feature detection, using coincidence detection (CD) mechanisms, made possible by the presence of…
▽ More
An increasing number of neuroscience studies are highlighting the importance of spatial dendritic branching in pyramidal neurons in the brain for supporting non-linear computation through localized synaptic integration. In particular, dendritic branches play a key role in temporal signal processing and feature detection, using coincidence detection (CD) mechanisms, made possible by the presence of synaptic delays that align temporally disparate inputs for effective integration. Computational studies on spiking neural networks further highlight the significance of delays for CD operations, enabling spatio-temporal pattern recognition within feed-forward neural networks without the need for recurrent architectures. In this work, we present DenRAM, the first realization of a spiking neural network with analog dendritic circuits, integrated into a 130nm technology node coupled with resistive memory (RRAM) technology. DenRAM's dendritic circuits use the RRAM devices to implement both delays and synaptic weights in the network. By configuring the RRAM devices to reproduce bio-realistic timescales, and through exploiting their heterogeneity, we experimentally demonstrate DenRAM's capability to replicate synaptic delay profiles, and efficiently implement CD for spatio-temporal pattern recognition. To validate the architecture, we conduct comprehensive system-level simulations on two representative temporal benchmarks, highlighting DenRAM's resilience to analog hardware noise, and its superior accuracy compared to recurrent architectures with an equivalent number of parameters. DenRAM not only brings rich temporal processing capabilities to neuromorphic architectures, but also reduces the memory footprint of edge devices, provides high accuracy on temporal benchmarks, and represents a significant step-forward in low-power real-time signal processing technologies.
△ Less
Submitted 14 December, 2023;
originally announced December 2023.
-
Synaptic metaplasticity with multi-level memristive devices
Authors:
Simone D'Agostino,
Filippo Moro,
Tifenn Hirtzlin,
Julien Arcamone,
Niccolò Castellani,
Damien Querlioz,
Melika Payvand,
Elisa Vianello
Abstract:
Deep learning has made remarkable progress in various tasks, surpassing human performance in some cases. However, one drawback of neural networks is catastrophic forgetting, where a network trained on one task forgets the solution when learning a new one. To address this issue, recent works have proposed solutions based on Binarized Neural Networks (BNNs) incorporating metaplasticity. In this work…
▽ More
Deep learning has made remarkable progress in various tasks, surpassing human performance in some cases. However, one drawback of neural networks is catastrophic forgetting, where a network trained on one task forgets the solution when learning a new one. To address this issue, recent works have proposed solutions based on Binarized Neural Networks (BNNs) incorporating metaplasticity. In this work, we extend this solution to quantized neural networks (QNNs) and present a memristor-based hardware solution for implementing metaplasticity during both inference and training. We propose a hardware architecture that integrates quantized weights in memristor devices programmed in an analog multi-level fashion with a digital processing unit for high-precision metaplastic storage. We validated our approach using a combined software framework and memristor based crossbar array for in-memory computing fabricated in 130 nm CMOS technology. Our experimental results show that a two-layer perceptron achieves 97% and 86% accuracy on consecutive training of MNIST and Fashion-MNIST, equal to software baseline. This result demonstrates immunity to catastrophic forgetting and the resilience to analog device imperfections of the proposed solution. Moreover, our architecture is compatible with the memristor limited endurance and has a 15x reduction in memory
△ Less
Submitted 21 June, 2023;
originally announced June 2023.
-
Dendritic Computation through Exploiting Resistive Memory as both Delays and Weights
Authors:
Melika Payvand,
Simone D'Agostino,
Filippo Moro,
Yigit Demirag,
Giacomo Indiveri,
Elisa Vianello
Abstract:
Biological neurons can detect complex spatio-temporal features in spiking patterns via their synapses spread across across their dendritic branches. This is achieved by modulating the efficacy of the individual synapses, and by exploiting the temporal delays of their response to input spikes, depending on their position on the dendrite. Inspired by this mechanism, we propose a neuromorphic hardwar…
▽ More
Biological neurons can detect complex spatio-temporal features in spiking patterns via their synapses spread across across their dendritic branches. This is achieved by modulating the efficacy of the individual synapses, and by exploiting the temporal delays of their response to input spikes, depending on their position on the dendrite. Inspired by this mechanism, we propose a neuromorphic hardware architecture equipped with multiscale dendrites, each of which has synapses with tunable weight and delay elements. Weights and delays are both implemented using Resistive Random Access Memory (RRAM). We exploit the variability in the high resistance state of RRAM to implement a distribution of delays in the millisecond range for enabling spatio-temporal detection of sensory signals. We demonstrate the validity of the approach followed with a RRAM-aware simulation of a heartbeat anomaly detection task. In particular we show that, by incorporating delays directly into the network, the network's power and memory footprint can be reduced by up to 100x compared to equivalent state-of-the-art spiking recurrent networks with no delays.
△ Less
Submitted 14 December, 2023; v1 submitted 11 May, 2023;
originally announced May 2023.
-
Clinical Deterioration Prediction in Brazilian Hospitals Based on Artificial Neural Networks and Tree Decision Models
Authors:
Hamed Yazdanpanah,
Augusto C. M. Silva,
Murilo Guedes,
Hugo M. P. Morales,
Leandro dos S. Coelho,
Fernando G. Moro
Abstract:
Early recognition of clinical deterioration (CD) has vital importance in patients' survival from exacerbation or death. Electronic health records (EHRs) data have been widely employed in Early Warning Scores (EWS) to measure CD risk in hospitalized patients. Recently, EHRs data have been utilized in Machine Learning (ML) models to predict mortality and CD. The ML models have shown superior perform…
▽ More
Early recognition of clinical deterioration (CD) has vital importance in patients' survival from exacerbation or death. Electronic health records (EHRs) data have been widely employed in Early Warning Scores (EWS) to measure CD risk in hospitalized patients. Recently, EHRs data have been utilized in Machine Learning (ML) models to predict mortality and CD. The ML models have shown superior performance in CD prediction compared to EWS. Since EHRs data are structured and tabular, conventional ML models are generally applied to them, and less effort is put into evaluating the artificial neural network's performance on EHRs data. Thus, in this article, an extremely boosted neural network (XBNet) is used to predict CD, and its performance is compared to eXtreme Gradient Boosting (XGBoost) and random forest (RF) models. For this purpose, 103,105 samples from thirteen Brazilian hospitals are used to generate the models. Moreover, the principal component analysis (PCA) is employed to verify whether it can improve the adopted models' performance. The performance of ML models and Modified Early Warning Score (MEWS), an EWS candidate, are evaluated in CD prediction regarding the accuracy, precision, recall, F1-score, and geometric mean (G-mean) metrics in a 10-fold cross-validation approach. According to the experiments, the XGBoost model obtained the best results in predicting CD among Brazilian hospitals' data.
△ Less
Submitted 17 December, 2022;
originally announced December 2022.
-
Hardware calibrated learning to compensate heterogeneity in analog RRAM-based Spiking Neural Networks
Authors:
Filippo Moro,
E. Esmanhotto,
T. Hirtzlin,
N. Castellani,
A. Trabelsi,
T. Dalgaty,
G. Molas,
F. Andrieu,
S. Brivio,
S. Spiga,
G. Indiveri,
M. Payvand,
E. Vianello
Abstract:
Spiking Neural Networks (SNNs) can unleash the full power of analog Resistive Random Access Memories (RRAMs) based circuits for low power signal processing. Their inherent computational sparsity naturally results in energy efficiency benefits. The main challenge implementing robust SNNs is the intrinsic variability (heterogeneity) of both analog CMOS circuits and RRAM technology. In this work, we…
▽ More
Spiking Neural Networks (SNNs) can unleash the full power of analog Resistive Random Access Memories (RRAMs) based circuits for low power signal processing. Their inherent computational sparsity naturally results in energy efficiency benefits. The main challenge implementing robust SNNs is the intrinsic variability (heterogeneity) of both analog CMOS circuits and RRAM technology. In this work, we assessed the performance and variability of RRAM-based neuromorphic circuits that were designed and fabricated using a 130\,nm technology node. Based on these results, we propose a Neuromorphic Hardware Calibrated (NHC) SNN, where the learning circuits are calibrated on the measured data. We show that by taking into account the measured heterogeneity characteristics in the off-chip learning phase, the NHC SNN self-corrects its hardware non-idealities and learns to solve benchmark tasks with high accuracy. This work demonstrates how to cope with the heterogeneity of neurons and synapses for increasing classification accuracy in temporal tasks.
△ Less
Submitted 10 February, 2022;
originally announced February 2022.
-
PCM-trace: Scalable Synaptic Eligibility Traces with Resistivity Drift of Phase-Change Materials
Authors:
Yigit Demirag,
Filippo Moro,
Thomas Dalgaty,
Gabriele Navarro,
Charlotte Frenkel,
Giacomo Indiveri,
Elisa Vianello,
Melika Payvand
Abstract:
Dedicated hardware implementations of spiking neural networks that combine the advantages of mixed-signal neuromorphic circuits with those of emerging memory technologies have the potential of enabling ultra-low power pervasive sensory processing. To endow these systems with additional flexibility and the ability to learn to solve specific tasks, it is important to develop appropriate on-chip lear…
▽ More
Dedicated hardware implementations of spiking neural networks that combine the advantages of mixed-signal neuromorphic circuits with those of emerging memory technologies have the potential of enabling ultra-low power pervasive sensory processing. To endow these systems with additional flexibility and the ability to learn to solve specific tasks, it is important to develop appropriate on-chip learning mechanisms.Recently, a new class of three-factor spike-based learning rules have been proposed that can solve the temporal credit assignment problem and approximate the error back-propagation algorithm on complex tasks. However, the efficient implementation of these rules on hybrid CMOS/memristive architectures is still an open challenge. Here we present a new neuromorphic building block,called PCM-trace, which exploits the drift behavior of phase-change materials to implement long lasting eligibility traces, a critical ingredient of three-factor learning rules. We demonstrate how the proposed approach improves the area efficiency by >10X compared to existing solutions and demonstrates a techno-logically plausible learning algorithm supported by experimental data from device measurements
△ Less
Submitted 16 February, 2021; v1 submitted 14 February, 2021;
originally announced February 2021.
-
US-net for robust and efficient nuclei instance segmentation
Authors:
Zhaoyang Xu,
Faranak Sobhani,
Carlos Fernandez Moro,
Qianni Zhang
Abstract:
We present a novel neural network architecture, US-Net, for robust nuclei instance segmentation in histopathology images. The proposed framework integrates the nuclei detection and segmentation networks by sharing their outputs through the same foundation network, and thus enhancing the performance of both. The detection network takes into account the high-level semantic cues with contextual infor…
▽ More
We present a novel neural network architecture, US-Net, for robust nuclei instance segmentation in histopathology images. The proposed framework integrates the nuclei detection and segmentation networks by sharing their outputs through the same foundation network, and thus enhancing the performance of both. The detection network takes into account the high-level semantic cues with contextual information, while the segmentation network focuses more on the low-level details like the edges. Extensive experiments reveal that our proposed framework can strengthen the performance of both branch networks in an integrated architecture and outperforms most of the state-of-the-art nuclei detection and segmentation networks.
△ Less
Submitted 31 January, 2019;
originally announced February 2019.
-
GAN-based Virtual Re-Staining: A Promising Solution for Whole Slide Image Analysis
Authors:
Zhaoyang Xu,
Xingru Huang,
Carlos Fernández Moro,
Béla Bozóky,
Qianni Zhang
Abstract:
Histopathological cancer diagnosis is based on visual examination of stained tissue slides. Hematoxylin and eosin (H\&E) is a standard stain routinely employed worldwide. It is easy to acquire and cost effective, but cells and tissue components show low-contrast with varying tones of dark blue and pink, which makes difficult visual assessments, digital image analysis, and quantifications. These li…
▽ More
Histopathological cancer diagnosis is based on visual examination of stained tissue slides. Hematoxylin and eosin (H\&E) is a standard stain routinely employed worldwide. It is easy to acquire and cost effective, but cells and tissue components show low-contrast with varying tones of dark blue and pink, which makes difficult visual assessments, digital image analysis, and quantifications. These limitations can be overcome by IHC staining of target proteins of the tissue slide. IHC provides a selective, high-contrast imaging of cells and tissue components, but their use is largely limited by a significantly more complex laboratory processing and high cost. We proposed a conditional CycleGAN (cCGAN) network to transform the H\&E stained images into IHC stained images, facilitating virtual IHC staining on the same slide. This data-driven method requires only a limited amount of labelled data but will generate pixel level segmentation results. The proposed cCGAN model improves the original network \cite{zhu_unpaired_2017} by adding category conditions and introducing two structural loss functions, which realize a multi-subdomain translation and improve the translation accuracy as well. % need to give reasons here. Experiments demonstrate that the proposed model outperforms the original method in unpaired image translation with multi-subdomains. We also explore the potential of unpaired images to image translation method applied on other histology images related tasks with different staining techniques.
△ Less
Submitted 8 July, 2022; v1 submitted 13 January, 2019;
originally announced January 2019.
-
A Passivity-based Concurrent Whole-Body Control (cWBC) of Persistently Interacting Human-Exoskeleton Systems
Authors:
Federico L. Moro,
Niccolò Iannacci,
Giovanni Legnani,
Lorenzo Molinari Tosatti
Abstract:
This paper presents a concurrent whole-body control (cWBC) for human-exoskeleton systems that are tightly coupled at a Cartesian level (e.g., feet, hands, torso). The exoskeleton generates joint torques that i) cancel the effects of gravity on the coupled system, ii) perform a primary task (e.g., maintaining the balance of the system), and iii) exploit the kinematic redundancy of the system to amp…
▽ More
This paper presents a concurrent whole-body control (cWBC) for human-exoskeleton systems that are tightly coupled at a Cartesian level (e.g., feet, hands, torso). The exoskeleton generates joint torques that i) cancel the effects of gravity on the coupled system, ii) perform a primary task (e.g., maintaining the balance of the system), and iii) exploit the kinematic redundancy of the system to amplify the forces exerted by the human operator. The coupled dynamic system is demonstrated to be passive, as its overall energy always goes dissipated until a minimum is reached. The proposed method is designed specifically to control exoskeletons for power augmentation worn by healthy operators in applications such as manufacturing, as it allows to increase the worker's capabilities, therefore reducing the risk of injuries.
△ Less
Submitted 9 August, 2017;
originally announced August 2017.
-
An Insight on the Ratio of Transmission of Motion (RoToM) and its Relation to the Centroidal Inertia Matrix
Authors:
Federico L. Moro
Abstract:
This paper analyses the dynamic response of a robot when subject to an external force that is applied to its Center of Mass (CoM). The Ratio of Transmission of Motion (RoToM) is proposed as a novel indicator of what part of the applied force generates motion, and what part is dissipated by the passive forces due to mechanical constraints. It depends on the configuration of the robot and on the dir…
▽ More
This paper analyses the dynamic response of a robot when subject to an external force that is applied to its Center of Mass (CoM). The Ratio of Transmission of Motion (RoToM) is proposed as a novel indicator of what part of the applied force generates motion, and what part is dissipated by the passive forces due to mechanical constraints. It depends on the configuration of the robot and on the direction of the force, and is always between 0 and 1. Extending this concept, a transmissibility ellipsoid is used to describe the behavior of the robot given a certain configuration, and varying the direction of the applied force. Another physical measure that is related to the transmissibility ellipsoid is the transmissibility index: it provides an indication on how similarly the system behaves when subject to forces coming from different directions. The presented analysis aims to provide a deeper insight on the centroidal dynamics of a robot, and on its dependence on the configuration. It can be beneficial for develo** whole-body controllers of redundant robots for e.g., reducing the effort in terms of joint torques to compensate for gravity, and more in general for designing interaction control architectures.
△ Less
Submitted 7 August, 2017;
originally announced August 2017.
-
Follow, listen, feel and go: alternative guidance systems for a walking assistance device
Authors:
Federico Moro,
Daniele Fontanelli,
Roberto Passerone,
Domenico Prattichizzo,
Luca Rizzon,
Stefano Scheggi,
Stefano Targher,
Antonella De Angeli,
Luigi Palopoli
Abstract:
In this paper, we propose several solutions to guide an older adult along a safe path using a robotic walking assistant (the c-Walker). We consider four different possibilities to execute the task. One of them is mechanical, with the c-Walker playing an active role in setting the course. The other ones are based on tactile or acoustic stimuli, and suggest a direction of motion that the user is sup…
▽ More
In this paper, we propose several solutions to guide an older adult along a safe path using a robotic walking assistant (the c-Walker). We consider four different possibilities to execute the task. One of them is mechanical, with the c-Walker playing an active role in setting the course. The other ones are based on tactile or acoustic stimuli, and suggest a direction of motion that the user is supposed to take on her own will. We describe the technological basis for the hardware components implementing the different solutions, and show specialized path following algorithms for each of them. The paper reports an extensive user validation activity with a quantitative and qualitative analysis of the different solutions. In this work, we test our system just with young participants to establish a safer methodology that will be used in future studies with older adults.
△ Less
Submitted 15 January, 2016;
originally announced January 2016.
