Research map / Label-efficient and robust learning / Recognition and long-tail
Learning Context-Conditioned Predicate Semantics via Prototype Feedback
NamGyu Jung, Chang Choi
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 changed | Where | Measured | Effect |
|---|---|---|---|
| add edge update to PE-Net (ablation) | PE-Net · VG-150 PredCls | mR@100: +2.3 | Improved |
| add edge update to PE-Net (ablation) | PE-Net · VG-150 SGDet | mR@100: +0.6 | Improved |
| add concatenation-based prototype alignment to edge-only variant (ablation) | AlignG · VG-150 SGCls | mR@100: +2.2 | Improved |
| replace concatenation prototype update with GRU (ablation) | AlignG · VG-150 PredCls | mR@100: +1.2 | Improved |
| replace concatenation prototype update with GRU (ablation) | AlignG · VG-150 SGCls | mR@100: +0.4 | Improved |
| replace concatenation prototype update with GRU (ablation) | AlignG · VG-150 SGDet | mR@100: +0.5 | Improved |
| use GRU update instead of EMA (ablation) | AlignG · VG-150 PredCls | F@100: 43.75 -> 47.47 | Improved |
| use GRU update instead of identity (no update) (ablation) | AlignG · VG-150 PredCls | R@100: 69.06 -> 65.06 | Hurt |
| add similarity penalty (ablation) | AlignG · VG-150 PredCls | F@100: 41.5 -> 47.5 | Improved |
| add diversity constraint (ablation) | AlignG · VG-150 PredCls | F@100: 38.6 -> 47.5 | Improved |
| add alignment loss (ablation) | AlignG · VG-150 PredCls | F@100: 42.7 -> 47.5 | Improved |
| double all loss weights (ablation) | AlignG · VG-150 PredCls | F@100: 47.5 -> 38.7 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- RA-SGG — RA-SGG: Retrieval-Augmented Scene Graph Generation Framework via Multi-Prototype Learning (AAAI 2025) · better on 7 of 7 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Panoptic Scene Graph Generation with Semantics-Prototype Learning (AAAI 2024 · similarity 0.89)
- Leveraging Predicate and Triplet Learning for Scene Graph Generation (CVPR 2024 · similarity 0.88)
- Multi-Prototype Space Learning for Commonsense-Based Scene Graph Generation (AAAI 2024 · similarity 0.88)
- Noise-Guided Predicate Representation Extraction and Diffusion-Enhanced Discretization for Scene Graph Generation (ICML 2025 · similarity 0.88)
- LLM4SGG: Large Language Models for Weakly Supervised Scene Graph Generation (CVPR 2024 · similarity 0.88)
- Scene Graph Thinking: Reinforcing Structured Visual Reasoning for Multimodal Large Language Models (ICML 2026 · similarity 0.88)
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.
