AI Paper Atlas

Research map / Architectures and efficiency / Attention and transformers

Super Vision Transformer

Mingbao Lin, Mengzhao Chen, Yuxin Zhang, Chunhua Shen, Rongrong Ji, Liujuan Cao

IJCV 2023 · Architectures and efficiency › Attention and transformers · approach: Efficient and robust vision transformers · 40 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
Train one ViT with multi-size patches and multiple keeping rates (SuperViT)DeiT-S (14x14, keep 1.0) · ImageNet-1k valTop-1 accuracy (%): 79.8 -> 80.6Improved
Apply SuperViT at 14x14 and keep 0.5 (half the FLOPs)DeiT-S · ImageNet-1k valTop-1 accuracy (%): 79.8 -> 80.0Improved
Apply SuperViT to larger LV-ViT-S backboneLV-ViT-S (14x14, keep 1.0) · ImageNet-1k valTop-1 accuracy (%): 83.3 -> 83.5Improved
Use SuperViT's 8x8 keep 0.5 subnet instead of DeiT-TDeiT-T · ImageNet-1k valTop-1 accuracy (%): 72.2 -> 73.9Improved
Use SuperViT instead of EViT token dropping at same FLOPsDeiT-S (3.0G FLOPs) · ImageNet-1k valTop-1 accuracy (%): 79.5 -> 80.5Improved
Train each subnet individually from scratch vs shared SuperViT training (adding SuperViT) (ablation)DeiT-S · ImageNet-1k valTop-1 accuracy (%): 78.0% -> 80.0%Improved
Add SuperViT training with only four subnets per iteration vs training all 12 (ablation)DeiT-S · ImageNet-1ktraining days: 8.8 -> 4.1 daysImproved
Fine-tune from SuperViT instead of DeiT-S on CIFAR-100DeiT-S · CIFAR-100Top-1 accuracy (%): 88.55 -> 89.33Improved
Use SuperViT pretrained backbone for UperNet segmentationDeiT-S · ADE20KmIoU: 42.96 -> 45.32Improved
Drop only low-attended tokens instead of also dropping 20% high-attended tokens (ablation)SuperViT DeiT-S 14x14 keep 0.5 · ImageNet-1k valTop-1 accuracy (%): 78.9 -> 80.0Improved
Drop tokens at blocks 4/7/10 instead of 4/6/8 (ablation)SuperViT DeiT-S 8x8 keep 1.0 · ImageNet-1k valTop-1 accuracy (%): 74.9 -> 75.8Improved
Drop tokens at blocks 4/7/10 instead of 3/6/9 (ablation)SuperViT DeiT-S 8x8 keep 1.0 · ImageNet-1k valTop-1 accuracy (%): 75.5 -> 75.8Improved

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