Compressing Deep Neural Networks Using Toeplitz Matrix: Algorithm Design and Fpga Implementation

Siyu Liao; Ashkan Samiee; Chunhua Deng; Yu Bai; Bo Yuan

ICASSP 2019accepted0 citations

Compressing Deep Neural Networks Using Toeplitz Matrix: Algorithm Design and Fpga Implementation

Siyu Liao, Ashkan Samiee, Chunhua Deng, Yu Bai, Bo Yuan

Abstract

Deep neural networks (DNNs) have emerged as an important artificial intelligence technique. However, the computation-intensive and storage-intensive DNNs pose severe challenges on efficient execution over the underlying hardware platform. In this paper we propose to impose Toeplitz structure on DNN models to achieve high compression ratio with negligible performance loss. Accordingly, the hardware performance can be significantly improved after performing model compression. We evaluate the proposed approach on speech recognition and implement the corresponding compressed model on FPGA. Experimental results show that our approach enables high hardware performance while retaining high task performance.

BibTeX

@inproceedings{icassp2019_compressingdeepn,
  title = {Compressing Deep Neural Networks Using Toeplitz Matrix: Algorithm Design and Fpga Implementation},
  author = {Siyu Liao and Ashkan Samiee and Chunhua Deng and Yu Bai and Bo Yuan},
  booktitle = {ICASSP 2019},
  year = {2019}
}