Weight-Sharing-Papers

Weight Sharing Paper list for Deep Neural Network Compression

Weight Sharing

• ACDC: Weight Sharing in Atom-Coefficient Decomposed Convolution, 2020 [Paper]
• Learning Shared Filter Bases for Efficient ConvNets, 2020 [Paper]
• Neural Epitome Search for Architecture-Agnostic Network Compression, ICLR 2020 [Paper]
• FSNet: Compression of Deep Convolutional Neural Networks by Filter Summary, ICLR 2020 [Paper]
• Learning Filter Basis for Convolutional Neural Network Compression, ICCV 2019 [Paper] [Code]
• Dynamic Recursive Neural Network, CVPR 2019 [Paper]
• Learning Implicitly Recurrent CNNs Through Parameter Sharing, ICLR 2019 [Paper] [Code]

Weight Clustering

• Deep k-Means: Re-Training and Parameter Sharing with Harder Cluster Assignments for Compressing Deep Convolutions, ICML 2018 [Paper] [Code]
• Clustering Convolutional Kernels to Compress Deep Neural Networks, ECCV 2018 [Paper] [Code]

Weight Decomposition

• Stable Low-rank Tensor Decomposition for Compression of Convolutional Neural Network, ECCV 2020 [Paper]
• PENNI: Pruned Kernel Sharing for Efficient CNN Inference, ICML2020 [Paper] [Code]
• Speeding up Convolutional Neural Networks with Low Rank Expansions, BMVC 2014 [Paper]