Latency-aware Unified Dynamic Networks for Efficient Image Recognition

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Han, Yizeng, Liu, Zeyu, Yuan, Zhihang, Pu, Yifan, Wang, Chaofei, Song, Shiji, Huang, Gao
Format: Preprint
Published: 2023
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914685340614656
author Han, Yizeng
Liu, Zeyu
Yuan, Zhihang
Pu, Yifan
Wang, Chaofei
Song, Shiji
Huang, Gao
author_facet Han, Yizeng
Liu, Zeyu
Yuan, Zhihang
Pu, Yifan
Wang, Chaofei
Song, Shiji
Huang, Gao
contents Dynamic computation has emerged as a promising avenue to enhance the inference efficiency of deep networks. It allows selective activation of computational units, leading to a reduction in unnecessary computations for each input sample. However, the actual efficiency of these dynamic models can deviate from theoretical predictions. This mismatch arises from: 1) the lack of a unified approach due to fragmented research; 2) the focus on algorithm design over critical scheduling strategies, especially in CUDA-enabled GPU contexts; and 3) challenges in measuring practical latency, given that most libraries cater to static operations. Addressing these issues, we unveil the Latency-Aware Unified Dynamic Networks (LAUDNet), a framework that integrates three primary dynamic paradigms-spatially adaptive computation, dynamic layer skipping, and dynamic channel skipping. To bridge the theoretical and practical efficiency gap, LAUDNet merges algorithmic design with scheduling optimization, guided by a latency predictor that accurately gauges dynamic operator latency. We've tested LAUDNet across multiple vision tasks, demonstrating its capacity to notably reduce the latency of models like ResNet-101 by over 50% on platforms such as V100, RTX3090, and TX2 GPUs. Notably, LAUDNet stands out in balancing accuracy and efficiency. Code is available at: https://www.github.com/LeapLabTHU/LAUDNet.
format Preprint
id arxiv_https___arxiv_org_abs_2308_15949
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Latency-aware Unified Dynamic Networks for Efficient Image Recognition
Han, Yizeng
Liu, Zeyu
Yuan, Zhihang
Pu, Yifan
Wang, Chaofei
Song, Shiji
Huang, Gao
Computer Vision and Pattern Recognition
Dynamic computation has emerged as a promising avenue to enhance the inference efficiency of deep networks. It allows selective activation of computational units, leading to a reduction in unnecessary computations for each input sample. However, the actual efficiency of these dynamic models can deviate from theoretical predictions. This mismatch arises from: 1) the lack of a unified approach due to fragmented research; 2) the focus on algorithm design over critical scheduling strategies, especially in CUDA-enabled GPU contexts; and 3) challenges in measuring practical latency, given that most libraries cater to static operations. Addressing these issues, we unveil the Latency-Aware Unified Dynamic Networks (LAUDNet), a framework that integrates three primary dynamic paradigms-spatially adaptive computation, dynamic layer skipping, and dynamic channel skipping. To bridge the theoretical and practical efficiency gap, LAUDNet merges algorithmic design with scheduling optimization, guided by a latency predictor that accurately gauges dynamic operator latency. We've tested LAUDNet across multiple vision tasks, demonstrating its capacity to notably reduce the latency of models like ResNet-101 by over 50% on platforms such as V100, RTX3090, and TX2 GPUs. Notably, LAUDNet stands out in balancing accuracy and efficiency. Code is available at: https://www.github.com/LeapLabTHU/LAUDNet.
title Latency-aware Unified Dynamic Networks for Efficient Image Recognition
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2308.15949