AI Paper Atlas

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

Enhancing Representations Through Heterogeneous Self-Supervised Learning

Zhong-Yu Li, Bowen Yin, Yongxiang Liu, Li Liu, Ming‐Ming Cheng

TPAMI 2025 · Label-efficient and robust learning › Self-supervised and contrastive · approach: Data-efficient and robust self-supervision · 5 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 heterogeneous auxiliary head (HSSL)iBOT ViT-B/16 · ImageNet-1K linear probing (400 ep)top-1: 77.8 -> 79.4Improved
add HSSLDINO ViT-S/16 · ImageNet-1K k-NN (200 ep)top-1: 61.2 -> 65.5Improved
add HSSL (600 ep) vs longer baseline (1600 ep)iBOT ViT-B/16 · ImageNet-1K fine-tuningtop-1: 84.0 -> 84.1Improved
add HSSLiBOT ViT-B/16 · ADE20KmIoU: 47.9 -> 50.3Improved
use ConvNext auxiliary head instead of baseline (no head) (ablation)ViT base · ImageNet-1Ktop-1: 67.5 -> 72.7Improved
use ViT auxiliary head with ResNet base (similar architecture) (ablation)ResNet base · ImageNet-1Ktop-1: 63.2 -> 64.0Improved
add ConvNext head alongside ResMLP head (ablation)ViT base · ImageNet-1Ktop-1: 72.6 -> 73.7Improved
remove first shortcut in the auxiliary head (ablation)ViT + ConvNext head · ImageNet-1Ktop-1: 71.0 -> 72.7Improved
use auxiliary head to supervise base model (A->B) instead of base self-supervision (B->B) (ablation)HSSL ViT · ImageNet-1Ktop-1: 68.0 -> 72.7Improved
use MLP + token mixer head instead of MLP only (ablation)HSSL ViT · ImageNet-1Ktop-1: 70.0 -> 72.7Improved
use separate (non-shared) projection (ablation)HSSL ViT · ImageNet-1Ktop-1: 72.3 -> 72.7Improved
use serial instead of parallel connection of heads (ablation)HSSL ViT · ImageNet-1Ktop-1: 72.8 -> 73.9Improved

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