.png){kind=icon}
返回
Universal Deep Neural Network Compression
.png){kind=icon}
OA
AI
DOI：10.1109/JSTSP.2020.2975903.png)
.png){kind=icon}
.png){kind=icon}
分享.png){kind=icon}
收藏
摘要
En 中文
We consider compression of deep neural networks (DNNs) by weight quantization and lossless source coding for memory-efficient deployment. Whereas the previous work addressed non-universal scalar quantization and entropy source coding, we for the first time introduce universal DNN compression by universal vector quantization and universal source coding. In particular, the proposed scheme utilizes universal lattice quantization, which randomizes the source by uniform random dithering before lattice quantization and can perform near-optimally on any source without relying on knowledge of the source distribution. Moreover, we present a method of fine-tuning vector quantized DNNs to recover any accuracy loss due to quantization. From our experiments, we show that the proposed scheme compresses the MobileNet and ShuffleNet models trained on ImageNet with the state-of-the-art compression ratios of 10.7 and 8.8, respectively.
Keyword:
Entropy
Vector quantization
Lattices
Source coding
Image coding
Neural networks
Deep neural networks
lossy compression
universal compression
entropy coded vector quantization
universal quantization
AI总结
对已上传原文的论文进行重点信息的提取，主要内容包括：简要概述、研究摘要、背景介绍、关键亮点、图文解析、展望与总结。
期刊
IF:
13.7
论文数:
1.9K
被引数:
1.1W
机构
引用论文
.png){kind=icon}
分享收藏
查看更多内容