The search functionality is under construction.
The search functionality is under construction.

SENTEI: Filter-Wise Pruning with Distillation towards Efficient Sparse Convolutional Neural Network Accelerators

Masayuki SHIMODA, Youki SADA, Ryosuke KURAMOCHI, Shimpei SATO, Hiroki NAKAHARA

  • Full Text Views

    0

  • Cite this

Summary :

In the realization of convolutional neural networks (CNNs) in resource-constrained embedded hardware, the memory footprint of weights is one of the primary problems. Pruning techniques are often used to reduce the number of weights. However, the distribution of nonzero weights is highly skewed, which makes it more difficult to utilize the underlying parallelism. To address this problem, we present SENTEI*, filter-wise pruning with distillation, to realize hardware-aware network architecture with comparable accuracy. The filter-wise pruning eliminates weights such that each filter has the same number of nonzero weights, and retraining with distillation retains the accuracy. Further, we develop a zero-weight skipping inter-layer pipelined accelerator on an FPGA. The equalization enables inter-filter parallelism, where a processing block for a layer executes filters concurrently with straightforward architecture. Our evaluation of semantic-segmentation tasks indicates that the resulting mIoU only decreased by 0.4 points. Additionally, the speedup and power efficiency of our FPGA implementation were 33.2× and 87.9× higher than those of the mobile GPU. Therefore, our technique realizes hardware-aware network with comparable accuracy.

Publication
IEICE TRANSACTIONS on Information Vol.E103-D No.12 pp.2463-2470
Publication Date
2020/12/01
Publicized
2020/08/03
Online ISSN
1745-1361
DOI
10.1587/transinf.2020PAP0013
Type of Manuscript
Special Section PAPER (Special Section on Parallel, Distributed, and Reconfigurable Computing, and Networking)
Category
Computer System

Authors

Masayuki SHIMODA
  Tokyo Institute of Technology
Youki SADA
  Tokyo Institute of Technology
Ryosuke KURAMOCHI
  Tokyo Institute of Technology
Shimpei SATO
  Tokyo Institute of Technology
Hiroki NAKAHARA
  Tokyo Institute of Technology

Keyword