AI Paper Atlas

Research map / Architectures and efficiency / Attention and transformers

SPViT: Enabling Faster Vision Transformers via Latency-Aware Soft Token Pruning

Zhenglun Kong, Peiyan Dong, Xiaolong Ma, Xin Meng, Wei Niu, Mengshu Sun, Xuan Shen, Geng Yuan, Bin Ren, Hao Tang, Minghai Qin, Yanzhi Wang

ECCV 2022 · Architectures and efficiency › Attention and transformers · approach: Vision transformer pruning · 187 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
use learned token selector instead of random token removal (ablation)SPViT DeiT-T · ImageNet-1K at 0.90 GFLOPstop-1 accuracy (%): 69.87 -> 72.10Improved
use learned token selector instead of random token removal (ablation)SPViT DeiT-S · ImageNet-1K at 2.64 GFLOPstop-1 accuracy (%): 77.25 -> 79.34Improved
use token selector instead of structure (dimension) pruning (ablation)SPViT DeiT-T · ImageNet-1K at 0.90 GFLOPstop-1 accuracy (%): 70.32 -> 72.10Improved
use token selector instead of structure (dimension) pruning (ablation)SPViT DeiT-S · ImageNet-1K at 2.64 GFLOPstop-1 accuracy (%): 77.86 -> 79.34Improved
place first selector after block 3 instead of block 1 (ablation)SPViT DeiT-S · ImageNet-1Ktop-1 accuracy (%): 76.10 -> 79.34Improved
place last selector after block 9 instead of block 11 (ablation)SPViT DeiT-S · ImageNet-1Ktop-1 accuracy (%): 78.76 -> 79.34Improved
use three selector phases (3-6-9) instead of two (6-9) (ablation)SPViT DeiT-S · ImageNet-1Ktop-1 accuracy (%): 78.53 -> 79.34Improved
use four selectors (3-5-7-9) instead of three (3-6-9) (ablation)SPViT DeiT-S · ImageNet-1Ktop-1 accuracy (%): 79.34 -> 79.34No clear effect
apply SPViT token pruning to DeiT-TDeiT-T · ImageNet-1Ktop-1 accuracy (%): 72.20 -> 72.10; GFLOPs 31% lowerNo clear effect
apply SPViT to Swin-TSwin-T · ImageNet-1Ktop-1 accuracy (%): 81.20 -> 80.70; GFLOPs 4.50 -> 3.47Mixed
apply SPViT on mobile CPUDeiT-S · Samsung Galaxy S20 latencylatency (ms): 113 -> 60 msImproved
apply SPViT on FPGADeiT-T · Xilinx ZCU102 latencylatency (ms): 8.81 -> 5.60 msImproved

Later papers that compared with this method

2 later accepted papers report SPViT as a baseline in their results tables; 2 of them report a better number on most of the results they share with it.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy