-
Distilling particle knowledge for fast reconstruction at high-energy physics experiments
Authors:
Aritra Bal,
Tristan Brandes,
Fabio Iemmi,
Markus Klute,
Benedikt Maier,
Vinicius Mikuni,
Thea Aarrestad
Abstract:
Knowledge distillation is a form of model compression that allows artificial neural networks of different sizes to learn from one another. Its main application is the compactification of large deep neural networks to free up computational resources, in particular on edge devices. In this article, we consider proton-proton collisions at the High-Luminosity LHC (HL-LHC) and demonstrate a successful…
▽ More
Knowledge distillation is a form of model compression that allows artificial neural networks of different sizes to learn from one another. Its main application is the compactification of large deep neural networks to free up computational resources, in particular on edge devices. In this article, we consider proton-proton collisions at the High-Luminosity LHC (HL-LHC) and demonstrate a successful knowledge transfer from an event-level graph neural network (GNN) to a particle-level small deep neural network (DNN). Our algorithm, DistillNet, is a DNN that is trained to learn about the provenance of particles, as provided by the soft labels that are the GNN outputs, to predict whether or not a particle originates from the primary interaction vertex. The results indicate that for this problem, which is one of the main challenges at the HL-LHC, there is minimal loss during the transfer of knowledge to the small student network, while improving significantly the computational resource needs compared to the teacher. This is demonstrated for the distilled student network on a CPU, as well as for a quantized and pruned student network deployed on a field-programmable gate array. Our study proves that knowledge transfer between networks of different complexity can be used for fast artificial intelligence (AI) in high-energy physics that improves the expressiveness of observables over non-AI-based reconstruction algorithms. Such an approach can become essential at the HL-LHC experiments, e.g., to comply with the resource budget of their trigger stages.
△ Less
Submitted 7 May, 2024; v1 submitted 21 November, 2023;
originally announced November 2023.
-
Unsupervised Deformable Ultrasound Image Registration and Its Application for Vessel Segmentation
Authors:
FNU Abhimanyu,
Andrew L. Orekhov,
Ananya Bal,
John Galeotti,
Howie Choset
Abstract:
This paper presents a deep-learning model for deformable registration of ultrasound images at online rates, which we call U-RAFT. As its name suggests, U-RAFT is based on RAFT, a convolutional neural network for estimating optical flow. U-RAFT, however, can be trained in an unsupervised manner and can generate synthetic images for training vessel segmentation models. We propose and compare the reg…
▽ More
This paper presents a deep-learning model for deformable registration of ultrasound images at online rates, which we call U-RAFT. As its name suggests, U-RAFT is based on RAFT, a convolutional neural network for estimating optical flow. U-RAFT, however, can be trained in an unsupervised manner and can generate synthetic images for training vessel segmentation models. We propose and compare the registration quality of different loss functions for training U-RAFT. We also show how our approach, together with a robot performing force-controlled scans, can be used to generate synthetic deformed images to significantly expand the size of a femoral vessel segmentation training dataset without the need for additional manual labeling. We validate our approach on both a silicone human tissue phantom as well as on in-vivo porcine images. We show that U-RAFT generates synthetic ultrasound images with 98% and 81% structural similarity index measure (SSIM) to the real ultrasound images for the phantom and porcine datasets, respectively. We also demonstrate that synthetic deformed images from U-RAFT can be used as a data augmentation technique for vessel segmentation models to improve intersection-over-union (IoU) segmentation performance
△ Less
Submitted 23 June, 2023;
originally announced June 2023.
-
Statistical Shape Analysis of Shape Graphs with Applications to Retinal Blood-Vessel Networks
Authors:
Aditi Basu Bal,
Xiaoyang Guo,
Tom Needham,
Anuj Srivastava
Abstract:
This paper provides theoretical and computational developments in statistical shape analysis of shape graphs, and demonstrates them using analysis of complex data from retinal blood-vessel (RBV) networks. The shape graphs are represented by a set of nodes and edges (planar articulated curves) connecting some of these nodes. The goals are to utilize shapes of edges and connectivities and locations…
▽ More
This paper provides theoretical and computational developments in statistical shape analysis of shape graphs, and demonstrates them using analysis of complex data from retinal blood-vessel (RBV) networks. The shape graphs are represented by a set of nodes and edges (planar articulated curves) connecting some of these nodes. The goals are to utilize shapes of edges and connectivities and locations of nodes to: (1) characterize full shapes, (2) quantify shape differences, and (3) model statistical variability. We develop a mathematical representation, elastic Riemannian shape metrics, and associated tools for such statistical analysis. Specifically, we derive tools for shape graph registration, geodesics, summaries, and shape modeling. Geodesics are convenient for visualizing optimal deformations, and PCA helps in dimension reduction and statistical modeling. One key challenge here is comparisons of shape graphs with vastly different complexities (in number of nodes and edges). This paper introduces a novel multi-scale representation of shape graphs to handle this challenge. Using the notions of (1) ``effective resistance" to cluster nodes and (2) elastic shape averaging of edge curves, one can reduce shape graph complexity while maintaining overall structures. This way, we can compare shape graphs by bringing them to similar complexity. We demonstrate these ideas on Retinal Blood Vessel (RBV) networks taken from the STARE and DRIVE databases.
△ Less
Submitted 28 November, 2022;
originally announced November 2022.
-
Adversarial Patch Camouflage against Aerial Detection
Authors:
Ajaya Adhikari,
Richard den Hollander,
Ioannis Tolios,
Michael van Bekkum,
Anneloes Bal,
Stijn Hendriks,
Maarten Kruithof,
Dennis Gross,
Nils Jansen,
Guillermo Pérez,
Kit Buurman,
Stephan Raaijmakers
Abstract:
Detection of military assets on the ground can be performed by applying deep learning-based object detectors on drone surveillance footage. The traditional way of hiding military assets from sight is camouflage, for example by using camouflage nets. However, large assets like planes or vessels are difficult to conceal by means of traditional camouflage nets. An alternative type of camouflage is th…
▽ More
Detection of military assets on the ground can be performed by applying deep learning-based object detectors on drone surveillance footage. The traditional way of hiding military assets from sight is camouflage, for example by using camouflage nets. However, large assets like planes or vessels are difficult to conceal by means of traditional camouflage nets. An alternative type of camouflage is the direct misleading of automatic object detectors. Recently, it has been observed that small adversarial changes applied to images of the object can produce erroneous output by deep learning-based detectors. In particular, adversarial attacks have been successfully demonstrated to prohibit person detections in images, requiring a patch with a specific pattern held up in front of the person, thereby essentially camouflaging the person for the detector. Research into this type of patch attacks is still limited and several questions related to the optimal patch configuration remain open.
This work makes two contributions. First, we apply patch-based adversarial attacks for the use case of unmanned aerial surveillance, where the patch is laid on top of large military assets, camouflaging them from automatic detectors running over the imagery. The patch can prevent automatic detection of the whole object while only covering a small part of it. Second, we perform several experiments with different patch configurations, varying their size, position, number and saliency. Our results show that adversarial patch attacks form a realistic alternative to traditional camouflage activities, and should therefore be considered in the automated analysis of aerial surveillance imagery.
△ Less
Submitted 31 August, 2020;
originally announced August 2020.
-
Statistical Shape Analysis of Brain Arterial Networks (BAN)
Authors:
Xiaoyang Guo,
Aditi Basu Bal,
Tom Needham,
Anuj Srivastava
Abstract:
Structures of brain arterial networks (BANs) - that are complex arrangements of individual arteries, their branching patterns, and inter-connectivities - play an important role in characterizing and understanding brain physiology. One would like tools for statistically analyzing the shapes of BANs, i.e. quantify shape differences, compare population of subjects, and study the effects of covariates…
▽ More
Structures of brain arterial networks (BANs) - that are complex arrangements of individual arteries, their branching patterns, and inter-connectivities - play an important role in characterizing and understanding brain physiology. One would like tools for statistically analyzing the shapes of BANs, i.e. quantify shape differences, compare population of subjects, and study the effects of covariates on these shapes. This paper mathematically represents and statistically analyzes BAN shapes as elastic shape graphs. Each elastic shape graph is made up of nodes that are connected by a number of 3D curves, and edges, with arbitrary shapes. We develop a mathematical representation, a Riemannian metric and other geometrical tools, such as computations of geodesics, means and covariances, and PCA for analyzing elastic graphs and BANs. This analysis is applied to BANs after separating them into four components -- top, bottom, left, and right. This framework is then used to generate shape summaries of BANs from 92 subjects, and to study the effects of age and gender on shapes of BAN components. We conclude that while gender effects require further investigation, the age has a clear, quantifiable effect on BAN shapes. Specifically, we find an increased variance in BAN shapes as age increases.
△ Less
Submitted 22 March, 2022; v1 submitted 7 July, 2020;
originally announced July 2020.
-
Waves in Thin Oceans on Oblate Neutron Stars
Authors:
Bart F. A. van Baal,
Frank R. N. Chambers,
Anna L. Watts
Abstract:
Waves in thin fluid layers are important in various stellar and planetary problems. Due to rapid rotation such systems will become oblate, with a latitudinal variation in the gravitational acceleration across the surface of the object. In the case of accreting neutron stars, rapid rotation could lead to a polar radius smaller than the equatorial radius by a factor $\sim 0.8$. We investigate how th…
▽ More
Waves in thin fluid layers are important in various stellar and planetary problems. Due to rapid rotation such systems will become oblate, with a latitudinal variation in the gravitational acceleration across the surface of the object. In the case of accreting neutron stars, rapid rotation could lead to a polar radius smaller than the equatorial radius by a factor $\sim 0.8$. We investigate how the oblateness and a changing gravitational acceleration affect different hydrodynamic modes that exist in such fluid layers through analytic approximations and numerical calculations. The wave vectors of $g$-modes and Yanai modes increase for more oblate systems compared to spherical counterparts, although the impact of variations in the changing gravitational acceleration is effectively negligible. We find that for increased oblateness, Kelvin modes show less equatorial confinement and little change in their wave vector. For $r$-modes, we find that for more oblate systems the wave vector decreases. The exact manner of these changes for the $r$-modes depends on the model for the gravitational acceleration across the surface.
△ Less
Submitted 11 June, 2020;
originally announced June 2020.
