Revisiting Dynamic Convolution via Matrix Decomposition

Yunsheng Li, Yinpeng Chen, Xiyang Dai, mengchen liu, Dongdong Chen, Ye Yu, Lu Yuan, Zicheng Liu, Mei Chen, Nuno Vasconcelos

Keywords: supervised representation learning, efficient network, dynamic network, matrix decomposition

Abstract: Recent research in dynamic convolution shows substantial performance boost for efficient CNNs, due to the adaptive aggregation of K static convolution kernels. It has two limitations: (a) it increases the number of convolutional weights by K-times, and (b) the joint optimization of dynamic attention and static convolution kernels is challenging. In this paper, we revisit it from a new perspective of matrix decomposition and reveal the key issue is that dynamic convolution applies dynamic attention over channel groups after projecting into a higher dimensional latent space. To address this issue, we propose dynamic channel fusion to replace dynamic attention over channel groups. Dynamic channel fusion not only enables significant dimension reduction of the latent space, but also mitigates the joint optimization difficulty. As a result, our method is easier to train and requires significantly fewer parameters without sacrificing accuracy. Source code is at https://github.com/liyunsheng13/dcd.

One-sentence Summary: Efficient network with dynamic matrix decomposition

Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics

Code: [ liyunsheng13/dcd](https://github.com/liyunsheng13/dcd)

Data: [ImageNet](https://paperswithcode.com/dataset/imagenet)

Community Implementations: [ 1 code implementation](https://www.catalyzex.com/paper/arxiv:2103.08756/code)