Research map / Detection and segmentation / Open-set and few-shot detection
Visual Relationship Detection with Low Rank Non-Negative Tensor Decomposition
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| model the triplet distribution with a rank-5 mixture instead of rank-1 independent distributions (ablation) | Visual relationship detection model · VG150 phrase classification, free k | Recall@50: 37.35 -> 42.71 | Helped |
| model the triplet distribution with a rank-5 mixture instead of rank-1 independent distributions (ablation) | Visual relationship detection model · VRD phrase classification, free k | Recall@50: 39.01 -> 40.47 | Helped |
| increase tensor rank from rank-5 to higher tested ranks (ablation) | Visual relationship detection model · VG150 and VRD phrase classification, free k | Recall@50: 42.71 -> 41.74; 40.47 -> 39.63 | Hurt |
| add a box-pair annotation-selection variable (ablation) | Rank-5 visual relationship detection model · VG150 phrase classification, free k | Recall@50: 42.71 -> 46.25 | Helped |
| add a box-pair annotation-selection variable (ablation) | Rank-5 visual relationship detection model · VRD phrase classification, free k | Recall@50: 40.47 -> 42.89 | Helped |
| add the triplet-frequency prior (ablation) | Rank-5 model with X_sel · VG150 phrase classification, free k | Recall@50: 46.25 -> 47.54 | Helped |
| add the triplet-frequency prior (ablation) | Rank-5 model with X_sel · VRD phrase classification, free k | Recall@50: 42.89 -> 44.02 | Helped |
| use the full model rather than RelDN | VG150 phrase classification · free k, R@100 | Recall@100: 50.8 -> 54.69 | Helped |
Closest papers by meaning
- Tensorize, Factorize and Regularize: Robust Visual Relationship Learning (CVPR 2018 · similarity 0.92)
- On Exploring Undetermined Relationships for Visual Relationship Detection (CVPR 2019 · similarity 0.89)
- Localize, Assemble, and Predicate: Contextual Object Proposal Embedding for Visual Relation Detection (AAAI 2020 · similarity 0.88)
- Groupwise Query Specialization and Quality-Aware Multi-Assignment for Transformer-based Visual Relationship Detection (CVPR 2024 · similarity 0.87)
- Leveraging Predicate and Triplet Learning for Scene Graph Generation (CVPR 2024 · similarity 0.86)
- EGTR: Extracting Graph from Transformer for Scene Graph Generation (CVPR 2024 · 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.