AI Paper Atlas

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

Investigating Self-Supervised Methods for Label-Efficient Learning

Srinivasa Rao Nandam, Sara Atito, Zhenhua Feng, Josef Kittler, Muhammed Awais

IJCV 2025 · Label-efficient and robust learning › Self-supervised and contrastive · approach: Self-supervision in supervised pipelines · 3 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 MIM pixel reconstruction to clustering (ablation)ViT-S clustering (ME-MAX) · ImageNet-1K 1-shottop-1 accuracy: 6.2 -> 15.7Improved
add MIM pixel reconstruction to contrastive learning (ablation)ViT-S contrastive · ImageNet-1K 1-shottop-1 accuracy: 2.3 -> 6.4Improved
use clustering instead of contrastive learning (ablation)ViT-S · ImageNet-1K 1-shottop-1 accuracy: 2.3 -> 6.2Improved
use MIM alone instead of clustering alone (ablation)ViT-S · ImageNet-1K 1-shottop-1 accuracy: 6.2 -> 8.7Improved
use ME-MAX instead of centring for collapse avoidance (ablation)ViT-S clustering · ImageNet-1K 1-shottop-1 accuracy: 5.3 -> 6.2Improved
use ME-MAX instead of sinkhorn for collapse avoidance (ablation)ViT-S clustering · ImageNet-1K 5-shottop-1 accuracy: 6.5 -> 29.6Improved
add patch-level clustering to class-level clustering (ablation)ViT-S clustering · ImageNet-1K 1-shottop-1 accuracy: 6.2 -> 7.9Improved
add patch-level clustering to class-level clustering (ablation)ViT-S clustering · ImageNet-1K 5-shottop-1 accuracy: 29.6 -> 35.0Improved
use MaskCluster instead of best prior SSL methodViT-S · ImageNet-1K 1 image per classtop-1 accuracy: 7.9 -> 22.4Improved
use MaskCluster instead of MSN for FPTrans initialisationFPTrans ViT-S/16 · Pascal 5i 1-shotmIoU: 49.7 -> 63.5Improved
use MaskCluster instead of iBoT for full ImageNet-1K fine-tuningViT-S/16 · ImageNet-1K fulltop-1 accuracy: 82.3 -> 82.1No clear effect
use MaskCluster instead of iBoT for Cars fine-tuningViT-S/16 · Carstop-1 accuracy: 94.0 -> 93.6Hurt

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