Dynamic Spatial Sparsification for Efficient Vision Transformers and Convolutional Neural Networks

计算机科学 计算 人工智能 变压器 失败 安全性令牌 卷积神经网络 特征(语言学) 模式识别(心理学) 算法 并行计算 量子力学 语言学 物理 哲学 计算机安全 电压
作者
Yongming Rao,Zuyan Liu,Wenliang Zhao,Jie Zhou,Jiwen Lu
出处
期刊:IEEE Transactions on Pattern Analysis and Machine Intelligence [IEEE Computer Society]
卷期号:45 (9): 10883-10897 被引量:24
标识
DOI:10.1109/tpami.2023.3263826
摘要

In this paper, we present a new approach for model acceleration by exploiting spatial sparsity in visual data. We observe that the final prediction in vision Transformers is only based on a subset of the most informative regions, which is sufficient for accurate image recognition. Based on this observation, we propose a dynamic token sparsification framework to prune redundant tokens progressively and dynamically based on the input to accelerate vision Transformers. Specifically, we devise a lightweight prediction module to estimate the importance of each token given the current features. The module is added to different layers to prune redundant tokens hierarchically. While the framework is inspired by our observation of the sparse attention in vision Transformers, we find that the idea of adaptive and asymmetric computation can be a general solution for accelerating various architectures. We extend our method to hierarchical models including CNNs and hierarchical vision Transformers as well as more complex dense prediction tasks. To handle structured feature maps, we formulate a generic dynamic spatial sparsification framework with progressive sparsification and asymmetric computation for different spatial locations. By applying lightweight fast paths to less informative features and expressive slow paths to important locations, we can maintain the complete structure of feature maps while significantly reducing the overall computations. Extensive experiments on diverse modern architectures and different visual tasks demonstrate the effectiveness of our proposed framework. By hierarchically pruning 66% of the input tokens, our method greatly reduces 31% ∼ 35% FLOPs and improves the throughput by over 40% while the drop of accuracy is within 0.5% for various vision Transformers. By introducing asymmetric computation, a similar acceleration can be achieved on modern CNNs and Swin Transformers. Moreover, our method achieves promising results on more complex tasks including semantic segmentation and object detection. Our results clearly demonstrate that dynamic spatial sparsification offers a new and more effective dimension for model acceleration. Code is available at https://github.com/raoyongming/DynamicViT.
最长约 10秒,即可获得该文献文件

科研通智能强力驱动
Strongly Powered by AbleSci AI
更新
PDF的下载单位、IP信息已删除 (2025-6-4)

科研通是完全免费的文献互助平台,具备全网最快的应助速度,最高的求助完成率。 对每一个文献求助,科研通都将尽心尽力,给求助人一个满意的交代。
实时播报
1秒前
songflower发布了新的文献求助10
1秒前
Billy完成签到,获得积分10
1秒前
aabbccwy完成签到,获得积分10
2秒前
2秒前
等太阳的人wyr完成签到,获得积分10
2秒前
2秒前
敏感代云完成签到,获得积分10
3秒前
teborlee完成签到,获得积分10
3秒前
郭俊秀完成签到 ,获得积分10
5秒前
daqisong发布了新的文献求助10
5秒前
5秒前
隐形曼青应助美好斓采纳,获得50
5秒前
juju发布了新的文献求助10
5秒前
秦琳昕完成签到,获得积分10
5秒前
翻羽完成签到,获得积分10
5秒前
6秒前
TheSilencer发布了新的文献求助10
6秒前
敏感人杰发布了新的文献求助10
7秒前
7秒前
wushuping完成签到,获得积分10
7秒前
36456657完成签到,获得积分0
7秒前
8秒前
10秒前
min20210429完成签到,获得积分10
10秒前
大方的羊青完成签到,获得积分10
10秒前
10秒前
温婉的松鼠完成签到,获得积分10
10秒前
gfdsh完成签到,获得积分10
11秒前
Jiangnj完成签到,获得积分10
11秒前
yile完成签到,获得积分10
11秒前
四叶草完成签到 ,获得积分10
11秒前
小虫发布了新的文献求助10
11秒前
任性的鸵鸟完成签到,获得积分10
11秒前
面向阳光完成签到,获得积分10
12秒前
壹号发布了新的文献求助10
12秒前
songflower完成签到,获得积分10
12秒前
12秒前
mumu发布了新的文献求助10
12秒前
joe_liu完成签到,获得积分10
13秒前
高分求助中
Picture Books with Same-sex Parented Families: Unintentional Censorship 1000
A new approach to the extrapolation of accelerated life test data 1000
ACSM’s Guidelines for Exercise Testing and Prescription, 12th edition 500
Nucleophilic substitution in azasydnone-modified dinitroanisoles 500
不知道标题是什么 500
Indomethacinのヒトにおける経皮吸収 400
Phylogenetic study of the order Polydesmida (Myriapoda: Diplopoda) 370
热门求助领域 (近24小时)
化学 材料科学 医学 生物 工程类 有机化学 生物化学 物理 内科学 纳米技术 计算机科学 化学工程 复合材料 遗传学 基因 物理化学 催化作用 冶金 细胞生物学 免疫学
热门帖子
关注 科研通微信公众号,转发送积分 3976918
求助须知:如何正确求助?哪些是违规求助? 3521120
关于积分的说明 11206103
捐赠科研通 3257952
什么是DOI,文献DOI怎么找? 1798932
邀请新用户注册赠送积分活动 878017
科研通“疑难数据库(出版商)”最低求助积分说明 806723