A Unifying Tensor View for Lightweight CNNs
2023 IEEE 15th International Conference on ASIC (ASICON)(2023)
摘要
Despite the decomposition of convolutional kernels for lightweight CNNs being
well studied, existing works that rely on tensor network diagrams or
hyperdimensional abstraction lack geometry intuition. This work devises a new
perspective by linking a 3D-reshaped kernel tensor to its various slice-wise
and rank-1 decompositions, permitting a straightforward connection between
various tensor approximations and efficient CNN modules. Specifically, it is
discovered that a pointwise-depthwise-pointwise (PDP) configuration constitutes
a viable construct for lightweight CNNs. Moreover, a novel link to the latest
ShiftNet is established, inspiring a first-ever shift layer pruning that
achieves nearly 50% compression with < 1% drop in accuracy for ShiftResNet.
更多查看译文
关键词
Convolutional Neural Network,Tensor Decomposition,Shift Layer,Pruning
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要