Research map / Detection and segmentation / Open-set and few-shot detection
ObjectRelator: Enabling Cross-View Object Relation Understanding Across Ego-Centric and Exo-Centric Perspectives
Yuqian Fu, Runze Wang, Bin Ren, Guolei Sun, Biao Gong, Yanwei Fu, Danda Pani Paudel, Xuanjing Huang, Luc Van Gool
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 MCFuse text-visual condition fusion (ablation) | PSALM · Ego-Exo4D Ego2Exo (Small) | IoU: 39.7 -> 43.2 | Improved |
| add MCFuse text-visual condition fusion (ablation) | PSALM · Ego-Exo4D Exo2Ego (Small) | IoU: 44.1 -> 47.4 | Improved |
| add XObjAlign cross-view consistency loss (ablation) | PSALM · Ego-Exo4D Ego2Exo (Small) | IoU: 39.7 -> 43.8 | Improved |
| add XObjAlign cross-view consistency loss (ablation) | PSALM · Ego-Exo4D Exo2Ego (Small) | IoU: 44.1 -> 48.3 | Improved |
| add MCFuse on top of XObjAlign (ablation) | ObjectRelator · Ego-Exo4D Ego2Exo (Small) | IoU: 43.8 -> 44.3 | Improved |
| add two-stage training (stage 1 trains MCFuse) to cross attention with learnable residual (ablation) | MCFuse · Ego-Exo4D Ego2Exo (Small) | IoU: 41.3 -> 43.2 | Improved |
| add learnable residual to cross attention (both stage 2 only) (ablation) | MCFuse · Ego-Exo4D Ego2Exo (Small) | IoU: 22.0 -> 41.3 | Improved |
| use learnable k_lea instead of fixed 0.5 (ablation) | MCFuse · Ego-Exo4D Ego2Exo (Small) | IoU: 41.7 -> 43.2 | Improved |
| use Euclidean instead of cosine distance in XObjAlign (ablation) | XObjAlign · Ego-Exo4D Ego2Exo (Small) | IoU: 42.5 -> 43.8 | Improved |
| raise L_XObj weight from 1 to 10 (ablation) | XObjAlign · Ego-Exo4D Ego2Exo (Small) | IoU: 43.8 -> 40.3 | Hurt |
| place MCFuse after XObjAlign instead of before (ablation) | ObjectRelator · Ego-Exo4D Ego2Exo (Small) | IoU: 43.6 -> 44.3 | Improved |
| train on joint Ego2Exo and Exo2Ego data instead of separate models | ObjectRelator · Ego-Exo4D Exo2Ego (Small) | IoU: 49.2 -> 50.6 | Improved |
Later papers that compared with this method
2 later accepted papers report ObjectRelator as a baseline in their results tables; 2 of them report a better number on most of the results they share with it.
- VGGT-Segmentor: Geometry-Enhanced Cross-View Segmentation (CVPR 2026 · better on 2 of 2 shared results)
- O-MaMa: Learning Object Mask Matching between Egocentric and Exocentric Views (ICCV 2025 · better on 1 of 1 shared result)
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.
- PSALM — PSALM: Pixelwise Segmentation with Large Multi-modal Model (ECCV 2024) · better on 1 of 1 shared result
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
- V^2-SAM: Marrying SAM2 with Multi-Prompt Experts for Cross-View Object Correspondence (CVPR 2026 · similarity 0.91)
- O-MaMa: Learning Object Mask Matching between Egocentric and Exocentric Views (ICCV 2025 · similarity 0.90)
- Robust Egocentric Referring Video Object Segmentation via Dual-Modal Causal Intervention (NeurIPS 2025 · similarity 0.87)
- Robust Ego-Exo Correspondence with Long-Term Memory (NeurIPS 2025 · similarity 0.87)
- Object-X: Learning to Reconstruct Multi-Modal 3D Object Representations (NeurIPS 2025 · similarity 0.85)
- Causal Tracing of Object Representations in Large Vision Language Models: Mechanistic Interpretability and Hallucination Mitigation (AAAI 2026 · similarity 0.85)
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.
