AI Paper Atlas

Research map / Label-efficient and robust learning / Self-supervised and contrastive

Unsupervised Representation Learning by Balanced Self Attention Matching

Daniel Shalam, Simon Korman

ECCV 2024 · Label-efficient and robust learning › Self-supervised and contrastive · approach: Invariance-based self-supervised learning · 0 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
add attention balancing (Sinkhorn balanced target matrix) (ablation)BAM · ImageNet linear probing (ViT-S/16, 150 epochs)accuracy: collapse -> 73.7Improved
use global normalization over the full nk x nk matrix (ablation)BAM · ImageNet linear probing (ViT-S/16, 150 epochs)accuracy: 60.7 -> 73.7Improved
add multi-crop (2 global + 6 local views) (ablation)BAM · ImageNet linear probing (ViT-S/16, 150 epochs)accuracy: 68.1 -> 73.7Improved
mask positive (augmentation) pairs in the similarity matrix (ablation)BAM · ImageNet linear probing (ViT-S/16, 150 epochs)accuracy: 72.6 -> 73.7Improved
add EMA student-teacher (ablation)BAM · ImageNet linear probing (ViT-S/16, 150 epochs)accuracy: 72.3 -> 73.7Improved
add ContrastiveCrop adapted to ViT (ablation)BAM · ImageNet linear probing (ViT-S/16, 150 epochs)accuracy: 73.5 -> 73.7Improved
use BAM instead of DINO for ImageNet-1% semi-supervised fine-tuning (ViT-S)ViT-S/16 · ImageNet 1% labelstop-1 accuracy: 58.7 -> 60.0Improved
BAM vs DINO under linear probing, ViT-S/16 at 300 epochsViT-S/16 · ImageNet linear probingtop-1 accuracy: 75.9 (DINO) vs 75.0 (BAM)Hurt
use CAFormer-M36 backbone instead of ViT-B/16BAM linear probing · ImageNet linear probingaccuracy: 77.5% -> 78.9%Improved
use ResNet-50 instead of ViT-S/16 backboneBAM linear probing · ImageNet linear probingaccuracy: 74.8% vs 75.0%No clear effect
increase video frames T from 7 to 28 in DAVIS-2017 propagationBAM vs DINO · DAVIS-2017 (J&F)_m(J&F)_m: 62.4 -> 64.0 (BAM); 62.3 -> 63.1 (DINO)Improved

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