Multi-Task Architecture with Attention for Imaging Atmospheric Cherenkov Telescope Data Analysis

Gamma-ray reconstruction from Cherenkov telescope data is multi-task by nature in astrophysics. The image recorded in the Cherenkov camera pixels relates to the type, energy, incoming direction and distance of a particle from a telescope observation. We propose γ-PhysNet, a physically inspired multi-task deep neural network for gamma/proton particle classification, and gamma energy and direction reconstruction. We compare its performance with single task networks on Monte Carlo simulated data and demonstrate the interest of reconstructing the impact point as an auxiliary task. We also show that γ-PhysNet outperforms a widespread analysis method for gamma-ray reconstruction. Finally, we study attention methods to solve relevant use cases. All the experiments are conducted in the context of single telescope analysis for the Cherenkov Telescope Array data

[1]  D. Nieto,et al.  Exploring deep learning as an event classification method for the Cherenkov Telescope Array , 2017, 1709.05889.

[2]  Patrick Lambert,et al.  Indexed Operations for Non-rectangular Lattices Applied to Convolutional Neural Networks , 2019, VISIGRAPP.

[3]  Vladlen Koltun,et al.  Multi-Task Learning as Multi-Objective Optimization , 2018, NeurIPS.

[4]  Marc Chaumont,et al.  PELICAN: deeP architecturE for the LIght Curve ANalysis , 2019, Astronomy & Astrophysics.

[5]  David Picard,et al.  2D/3D Pose Estimation and Action Recognition Using Multitask Deep Learning , 2018, 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition.

[6]  Zhao Chen,et al.  GradNorm: Gradient Normalization for Adaptive Loss Balancing in Deep Multitask Networks , 2017, ICML.

[7]  Zhongfei Zhang,et al.  Partially Shared Multi-task Convolutional Neural Network with Local Constraint for Face Attribute Learning , 2018, 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition.

[8]  Jian Sun,et al.  Deep Residual Learning for Image Recognition , 2015, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[9]  D. Nieto,et al.  Studying Deep Convolutional Neural Networks With Hexagonal Lattices for Imaging Atmospheric Cherenkov Telescope Event Reconstruction , 2019, Proceedings of 36th International Cosmic Ray Conference — PoS(ICRC2019).

[10]  T. Le Flour,et al.  The Cherenkov Telescope Array Large Size Telescope , 2013, 1307.4565.

[11]  Chongruo Wu,et al.  ResNeSt: Split-Attention Networks , 2020, 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW).

[12]  Mathieu de Naurois,et al.  A high performance likelihood reconstruction of γ-rays for imaging atmospheric Cherenkov telescopes , 2009, 0907.2610.

[13]  Xilin Chen,et al.  Object-Contextual Representations for Semantic Segmentation , 2019, ECCV.

[14]  H. Collaboration.  Resolving the Crab pulsar wind nebula at teraelectronvolt energies , 2019, Nature Astronomy.

[15]  A. Quirrenbach,et al.  A very-high-energy component deep in the γ-ray burst afterglow , 2019, Nature.

[16]  R. D. Parsons,et al.  Background rejection in atmospheric Cherenkov telescopes using recurrent convolutional neural networks , 2019, The European Physical Journal C.

[17]  Li Fei-Fei,et al.  Dynamic Task Prioritization for Multitask Learning , 2018, ECCV.

[18]  T. Lohse,et al.  Application of deep learning methods to analysis of imaging atmospheric Cherenkov telescopes data , 2018, Astroparticle Physics.

[19]  Edgar Simo-Serra,et al.  Let there be Color!: Joint End-to-end Learning of Global and Local Image Priors for Automatic Image Colorization with Simultaneous Classification , 2016 .

[20]  Yoshua Bengio,et al.  Neural Machine Translation by Jointly Learning to Align and Translate , 2014, ICLR.

[21]  Roberto Cipolla,et al.  Multi-task Learning Using Uncertainty to Weigh Losses for Scene Geometry and Semantics , 2017, 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition.

[22]  R. D. Parsons,et al.  HESS II Data Analysis with ImPACT , 2015, 1509.06322.

[23]  Dario Pavllo,et al.  3D Human Pose Estimation in Video With Temporal Convolutions and Semi-Supervised Training , 2018, 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[24]  Yong Jae Lee,et al.  Cross-Domain Self-Supervised Multi-task Feature Learning Using Synthetic Imagery , 2017, 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition.

[25]  Marc Chaumont,et al.  A CNN adapted to time series for the classification of Supernovae , 2019, Color Imaging: Displaying, Processing, Hardcopy, and Applications.

[26]  Edward J. Kim,et al.  Star-galaxy Classification Using Deep Convolutional Neural Networks , 2016, ArXiv.

[27]  Jian Sun,et al.  Identity Mappings in Deep Residual Networks , 2016, ECCV.

[28]  Heinrich J. Völk,et al.  Imaging very high energy gamma-ray telescopes , 2008, 0812.4198.

[29]  Rich Caruana,et al.  Multitask Learning , 1998, Encyclopedia of Machine Learning and Data Mining.

[30]  Lukasz Kaiser,et al.  Attention is All you Need , 2017, NIPS.

[31]  A. Hillas.  Cerenkov light images of EAS produced by primary gamma , 1985 .

[32]  Bo Wang,et al.  SAUNet: Shape Attentive U-Net for Interpretable Medical Image Segmentation , 2020, MICCAI.

[33]  Sebastian Ruder,et al.  An Overview of Multi-Task Learning in Deep Neural Networks , 2017, ArXiv.

[34]  A.Fiasson,et al.  Optimization of multivariate analysis for IACT stereoscopic systems , 2010, 1004.3375.

[35]  Sebastian Thrun,et al.  Is Learning The n-th Thing Any Easier Than Learning The First? , 1995, NIPS.

[36]  Ross B. Girshick,et al.  Mask R-CNN , 2017, 1703.06870.

[37]  Dustin Tran,et al.  Image Transformer , 2018, ICML.

[38]  Han Zhang,et al.  Self-Attention Generative Adversarial Networks , 2018, ICML.

[39]  Jimmy Ba,et al.  Adam: A Method for Stochastic Optimization , 2014, ICLR.

[40]  Qi Feng,et al.  The analysis of VERITAS muon images using convolutional neural networks , 2016, Proceedings of the International Astronomical Union.

[41]  Juan José Rodríguez-Vázquez,et al.  Extracting Gamma-Ray Information from Images with Convolutional Neural Network Methods on Simulated Cherenkov Telescope Array Data , 2018, ANNPR.

[42]  Gaël Varoquaux,et al.  Scikit-learn: Machine Learning in Python , 2011, J. Mach. Learn. Res..

[43]  Matthijs Douze,et al.  Fixing the train-test resolution discrepancy , 2019, NeurIPS.