Research map / Label-efficient and robust learning / Self-supervised and contrastive
CG-SSL: Concept-Guided Self-Supervised Learning
Sara Atito, Josef Kittler, Imran Razzak, Muhammad Awais
Read the paperCheck your idea against 137,217 papers
AI-extracted summary. Written by AI from the paper's abstract only. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Proposes CG-SSL, a self-supervised visual representation framework trained with a three-phase curriculum that moves from global impressions to distinct object-like concept components.
Problem
Self-supervised visual representation learning lacks structure and interpretability; the abstract motivates moving from a global scene impression to object-like components.
Approach
A concept-guided curriculum of three training phases that first captures a global impression and then refines it into distinct, object-like concept components, bringing structure and interpretability to SSL representations.
Closest papers by meaning
- Harnessing small projectors and multiple views for efficient vision pretraining (NeurIPS 2024 · similarity 0.88)
- Bridging the gap to real-world language-grounded visual concept learning (NeurIPS 2025 · similarity 0.88)
- FALCON: Fast Visual Concept Learning by Integrating Images, Linguistic descriptions, and Conceptual Relations (ICLR 2022 · similarity 0.88)
- Adaptive Visual Scene Understanding: Incremental Scene Graph Generation (NeurIPS 2024 · similarity 0.86)
- SC-Tune: Unleashing Self-Consistent Referential Comprehension in Large Vision Language Models (CVPR 2024 · similarity 0.86)
- Beyond Scalars: Concept-Based Alignment Analysis in Vision Transformers (NeurIPS 2025 · similarity 0.86)
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.
