WP1.PD.5

CAN WE DISTILL KNOWLEDGE FROM POWERFUL TEACHERS DIRECTLY?

Chengyao Qian, Munawar Hayat, Mehrtash Harandi, Monash University, Australia

Session:
WP1.PD: Machine Learning Theory and Algorithms for Image Processing I Poster

Track:
Applications of Machine Learning

Location:
Poster Area D

Presentation Time:
Wed, 11 Oct, 14:30 - 16:00 Malaysia Time (UTC +8)

Session Chair:
Thomas Richter, Fraunhofer HHI
View Manuscript
Presentation
Not logged in.
Discussion
Not logged in.
Resources
Session WP1.PD
WP1.PD.1: METAGRAD: ADAPTIVE GRADIENT QUANTIZATION WITH HYPERNETWORKS
Kaixin Xu, A*STAR, Singapore; Alina Hui Xiu Lee, National University of Singapore, Singapore; Ziyuan Zhao, Zhe Wang, Min Wu, A*STAR, Singapore; Weisi Lin, Nanyang Technological University, Singapore
WP1.PD.2: DFT-CAM: DISCRETE FOURIER TRANSFORM DRIVEN CLASS ACTIVATION MAP
Yangyang Wang, Filiz Bunyak, University of Missouri-Columbia, United States of America
WP1.PD.3: CROSS-LAYER PATCH ALIGNMENT AND INTRA-AND-INTER PATCH RELATIONS FOR KNOWLEDGE DISTILLATION
Yi Zhang, Xi'an Jiaotong University, China; Yingke Gao, China Academy of Space Technology, China; Haonan Zhang, Xinyu Lei, Longjun Liu, Xi'an Jiaotong University, China
WP1.PD.4: DESIGNING STRONG BASELINES FOR TERNARY NEURAL NETWORK QUANTIZATION THROUGH SUPPORT AND MASS EQUALIZATION
Edouard Yvinec, Arnaud Dapogny, Kevin Bailly, Datakalab, France
WP1.PD.5: CAN WE DISTILL KNOWLEDGE FROM POWERFUL TEACHERS DIRECTLY?
Chengyao Qian, Munawar Hayat, Mehrtash Harandi, Monash University, Australia
WP1.PD.6: FALSE CORRESPONDENCE REMOVAL VIA REVISITING SEMANTIC CONTEXT WITH POSITION-ATTENTIVE LEARNING
Ruiyuan Li, Zhaolin Xiao, Meng Zhang, Haonan Su, Haiyan Jin, Xi'an University of Technology, China
WP1.PD.7: LEARNING TO DRAW THROUGH A MULTI-STAGE ENVIRONMENT MODEL BASED REINFORCEMENT LEARNING
Ji Qiu, Peng Lu, Beijing University of Posts and Telecommunications, China; Xujun Peng, Amazon Alexa AI, United States of America
WP1.PD.8: DODGING THE DOUBLE DESCENT IN DEEP NEURAL NETWORKS
Victor Quetu, Enzo Tartaglione, LTCI, Télécom Paris, Institut Polytechnique de Paris, France
WP1.PD.9: LKBQ: PUSHING THE LIMIT OF POST-TRAINING QUANTIZATION TO EXTREME 1 BIT
Tianxiang Li, Tsinghua University, China; Bin Chen, Harbin Institute of Technology, China; QianWei Wang, Yujun Huang, ShuTao Xia, Tsinghua University, China
WP1.PD.10: INFERENCE ACCELERATION OF DEEP LEARNING CLASSIFIERS BASED ON RNN
Fekhr Eddine Keddous, University Paris Est Créteil, Laboratoire LISSI / Cyclope.ai, France; Nadiya Shvai, Cyclope.ai, France; Arcadi Llanza, Amir Nakib, University Paris Est Créteil, Laboratoire LISSI / Cyclope.ai, France
WP1.PD.12: A Gradient Boosting Approach for Training Convolutional and Deep Neural Networks
Seyedsaman Emami, Gonzalo Martinez-Munoz, Escuela Politécnica Superior
Contacts