Detailed Program

Tuesday, October 20, 2020

Keynote 1: Low-power processors for signal processing at the edge Chairs: Andreas Burg (EPFL) and Andy Wu (National Taiwan University) Luca Benini, ETH Zürich, University of Bologna ▸ Video Presentation: 4PM Lisbon / 8AM California / 11 PM Beijing
▸ Questions & Answers (Live): 4:45PM Lisbon / 8:45AM California / 11:45 PM Beijing
▸ See additional details

Session 1: Design Methods of Signal Processing Systems Chairs: John McAllister (Queen’s University Belfast) and Daniel Menard (UEB/INSA Rennes/IETR) ▸ Manipulation of the Fundamental Frequency Micro-Variations Using a Fully Parametric and Computationally Efficient Speech Model, João P. Silva, Marco A. Oliveira, Clara F. Cardoso and Aníbal J. Ferreira
▸ Dynamic Hyperdimensional Computing for Improving Accuracy-Energy Efficiency Trade-offs, Yu-Chuan Chuang, Cheng-Yang Chang and An-Yeu Wu
▸ Optimizing Halide for Digital Signal Processors, John Stratton, Jyothi Krishna Sreelatha, Rajiv Ravindran, Sachin Dake and Jeevitha Palanisamy
▸ Semantic Segmentation Via Enhancing Context Information By Fusing Multiple High-Level Features, Kuang-Chien Li, Ching-Te Chiu and Shan-Chien Hsiao
▸ Polynomial Approximation with Non-Uniform Segmentation for Bivariate Functions, Percelay Mathieu, Justine Bonnot, Florian Arrestier and Daniel Menard
▸ Approximate Buffers for Reducing Memory Requirements: Case study on SKA, Hugo Miomandre, Jean-François Nezan, Daniel Menard, Adam Campbell, Anthony Griffin, Seth Hall and Andrew Ensor
▸ Point cloud denoising using joint geometry/color graph wavelets, Muhammad Abeer Irfan and Enrico Magli
▸ Graph Coordination for Compact Representation of Regular Dataflow Structures, John McAllister and Michael Davis
▸ Optimization of Probabilistic Switching Models Based on a Two-Step Clustering Approach, Hafsa Iqbal, Damian Campo, Giulia Slavic, Lucio Marcenaro, David Martin Gomez and Carlo Regazzoni

Session 2: Software Implementation of Signal Processing Systems Chairs: Marco Mattavelli (EPFL) and Ching-Te Chiu (National Tsing Hua University) ▸ Memory-Latency-Accuracy Trade-offs for Continual Learning on a RISC-V Extreme-Edge Node, Leonardo Ravaglia, Manuele Rusci, Alessandro Capotondi, Francesco Conti, Lorenzo Pellegrini, Vincenzo Lomonaco, Davide Maltoni and Luca Benini
▸ Programming heterogeneous CPU-GPU systems by high level dataflow synthesis, Aurelien Bloch, Endri Bezati and Marco Mattavelli
▸ Efficient implementation of convolution and winograd on ASMP embedded multicore vector processor, Nicolas Leclaire, Stéphane Mancini, Claude Delnondedieu and Jean-Paul Henriques
▸ Pushing the Limits of Energy Efficiency for Non-Binary LDPC Decoders on GPUs and FPGAs, Srinivasan Subramaniyan, Oscar Ferraz, Ashuthosh M R, Santosh Krishna, Guohui Wang, Joseph R. Cavallaro, Vitor Silva, Gabriel Falcao and Madhura Purnaprajna

PhD Forum Chairs: Alexios Balatsoukas-Stimming (Eindhoven University of Technology) ▸ A Reconfigurable DNN Training Accelerator on FPGA, Jinming Lu, Jun Lin and Zhongfeng Wang
▸ On the Use of Audio Fingerprinting Features for Speech Enhancement with Generative Adversarial Network, Farnood Faraji, Yazid Attabi, Benoit Champagne and Wei-Ping Zhu
▸ Towards real time interception of Frequency Hopping signals, Corentin Lavaud, Robin Gerzaguet, Matthieu Gautier, Olivier Berder, Erwan Nogues and Stephane Molton
▸ Complexity Estimation for Load Balancing of 360-degree Intra Versatile Video Coding, Jose N. Filipe, Joao Carreira, Luis M. N. Tavora, Sergio M. M. Faria, Antonio Navarro and Pedro A. Amado Assuncao

Wednesday, October 21, 2020

Keynote 2: How to Evaluate Efficient Deep Neural Network Approaches Chairs: Andreas Burg (EPFL) and Andy Wu (National Taiwan University) Vivienne Sze, MIT ▸ Video Presentation: 4PM Lisbon / 8AM California / 11 PM Beijing
▸ Questions & Answers (Live): 4:45PM Lisbon / 8:45AM California / 11:45 PM Beijing
▸ See additional details

Session 3: Machine Learning for Signal Processing Chairs: Zhongfeng Wang (Nanjing University) and Wonyong Sung (National Taiwan University) ▸ GrateTile: Efficient Sparse Tensor Tiling for CNN Processing, Yu-Sheng Lin, Hung-Chang Lu, Yang-Bin Tsao, Yi-Ming Chi, Wei-Chao Chen and Shao-Yi Chien
▸ Neural Network Based Decision Fusion for Abnormality Detection via Molecular Communications, Sinem Solak and Menguc Oner
▸ Knowledge Distillation for Optimization of Quantized Deep Neural Networks, Sungho Shin, Yoonho Boo and Wonyong Sung
▸ Quantized Neural Networks: Characterization and Holistic Optimization, Yoonho Boo, Sungho Shin and Wonyong Sung
▸ Temporal Residual Feature Learning for Efficient 3D Convolutional Neural Network on Action Recognition Task, Haonan Wang, Yuchen Mei, Jun Lin and Zhongfeng Wang
▸ Arithmetic Precision Reconfigurable Convolution Neural Network Accelerator, En-Ho Shen, Jan P. Klopp and Shao-Yi Chien
▸ Neural Network-Aided BCJR Algorithm for Joint Symbol Detection and Channel Decoding, Wen-Chiao Tsai, Chieh-Fang Teng, Han-Mo Ou and An-Yeu Wu
▸ Machine Learning Based Outdoor Localization Using the RSSI of Multibeam Antennas, Ning Xu, Sitian Li, Andreas Burg and Adrian Schumacher
▸ IP-HDC: Information-Preserved Hyperdimensional Computing for Multi-task Learning, Cheng-Yang Chang, Yu-Chuan Chuang and An-Yeu Wu

Session 4: Hardware Implementation of Signal Processing Systems Chairs: Xinmiao Zhang (The Ohio State University) and Christophe Jego (IMS Laboratory) ▸ Dynamic Fused Multiply-Accumulate Posit Unit With Variable Exponent Size For Low-Precision DSP Applications, Nuno Neves, Pedro Tomas and Nuno Roma
▸ Low-Complexity Architectures for Parallel Long BCH Encoders, Yok Jye Tang and Xinmiao Zhang
▸ Implementation of a High-Throughput Fast-SSC Polar Decoder with Sequence Repetition Node, Haotian Zheng, Alexios Balatsoukas-Stimming, Zizheng Cao and Ton Koonen
▸ A Modified Serial Commutator Architecture for Real-Valued Fast Fourier Transform, Sungjin Park and Dongsuk Jeon
▸ Low-latency Sorter Architecture for Polar Codes Successive-Cancellation-List Decoding, Bertrand Le Gal, Yann Delomier, Camille Leroux and Christophe Jego
▸ Transformer PUF: A Highly Flexible Configurable RO PUF Based on FPGA, Ziwei Wei, Yijun Cui, Yunpeng Chen, Chenghua Wang, Chongyan Gu and Weiqiang Liu
▸ Energy-Efficient Target Recognition using ReRAM Crossbars for Enabling On-Device Intelligence, Sourav Sanyal, Aayush Ankit, Craig M. Vineyard and Kaushik Roy

Thursday, October 22, 2020

Keynote 3: Machine Learning & AI for 5G systems Chairs: Andreas Burg (EPFL) and Andy Wu (National Taiwan University) Shilpa Talwar, Intel ▸ Video Presentation: 4PM Lisbon / 8AM California / 11 PM Beijing
▸ Questions & Answers (Live): 4:45PM Lisbon / 8:45AM California / 11:45 PM Beijing
▸ See additional details

Special Session 1: DSP in Communication Systems Chairs: Philippe Coussy (Universit de Bretagne Sud / Lab-STICC) and Oner Orhan (Intel) ▸ Machine Learning in Optical Communication Systems and Networks, Rui Morais and Bruno Pereira
▸ Experimental Investigation of Deep Learning for Digital Signal Processing in Short Reach Optical Fiber Communications, Boris Karanov, Mathieu Chagnon, Vahid Aref, Filipe Ferreira, Domanic Lavery, Polina Bayvel and Laurent Schmalen
▸ Principles and Techniques to Optimize Digital Communication Systems from End to End, Mathieu Chagnon
▸ Reducing Polar Decoding Latency by Neural Network-based On-the-fly Decoder Selection, Ahmed Elkelesh, Sebastian Cammerer and Stephan ten Brink
▸ Time-Domain Learned Digital Back-Propagation, Eric Sillekens, Polina Bayvel and Robert Killey
▸ Deep Learning of the Physical Layer for BICM Systems, Fayçal Ait Aoudia, Sebastian Cammerer, Sebastian Dörner, Maximilian Stark, Jakob Hoydis and Stephan ten Brink

Special Session 2: Architectures for Digital Signal Processing at the Edge Chairs: Philippe Coussy (Universit de Bretagne Sud / Lab-STICC) and Richard Dorrance (Intel) ▸ High-Throughput VLSI Architecture for GRAND, Syed Mohsin Abbas, Thibaud Tonnellier, Furkan Ercan and Warren J. Gross
▸ Deep Neural Networks Characterization Framework for Efficient Implementation on Embedded Systems, Nermine Ali, Jean-Marc Philippe, Benoit Tain, Thomas Peyret and Philippe Coussy
▸ Accelerating Computer Vision Algorithms on Heterogeneous Edge Computing Platforms, Alok Prakash, Nirmala Ramakrishnan, Kratika Garg and Thambipillai Srikanthan
▸ Fast and Efficient Decision-based Attack for Deep Neural Network on Edge, Himanshu Jain, Sakshi Rathore, Abdul Rahoof, Vivek Chaturvedi and Satyajit Das

Session 5: Signal Processing Application Systems Chairs: Vasuki Narasimha Swamy (Intel) and Warren Gross (McGill University) ▸ An Early-Stopping Mechanism for DSCF Decoding of Polar Codes, Ilshat Sagitov and Pascal Giard
▸ iGYM: An Inclusive Augmented Reality Exergame for People of All Abilities, Pengqi Lu, Jaechan Lim, Roland Graf and Hun-Seok Kim
▸ Acoustics Based Proximity Detector For Mobile Phones, Zacharie Cohen, Andy Rodan, Alon Eilam and Pavel Lifshits
▸ Hybrid Combining Design for User Clustering in mmWave MIMO Systems, Darian Pérez-Adán, Óscar Fresnedo, José P. González-Coma and Luis Castedo
▸ GPU-Based Linearization of MIMO Arrays, Chance Tarver, Arav Singhal and Joseph Cavallaro
▸ Financial Time Series Forecasting Model Based on EMD and Rolling Grey Model, Binjing Li, Fenggui Liu, Jun Lin and Zhongfeng Wang
▸ Real-Time Estimation of Direction of Arrival of Speech Source Using Three Microphones, Serkan Tokgoz, Anton Kovalyov and Issa Panahi
▸ End-to-End Learning of Video Compression Using Spatio-Temporal Autoencoders, Jorge Pessoa, Helena Aidos, Pedro Tomas and Mario A. T. Figueiredo
▸ ECG-aided PPG Signal Quality Assessment (SQA) System for Heart Rate Estimation, Yu-Chia Yang, Win-Ken Beh, Yi-Cheng Lo, An-Yeu (Andy) Wu, and Shih-Jen Lu