Research map / Vision-language and multimodal / Human and embodied modeling
Relational Visual Similarity
Thao Nguyen, Sicheng Mo, Krishna Kumar Singh, Yilin Wang, Jing Shi, Nicholas Kolkin, Eli Shechtman, Yong Jae Lee, Yuheng Li
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 |
|---|---|---|---|
| use anonymous captions from an image group instead of from a single image (ablation) | relsim retrieval · Relational retrieval | relational similarity score: 5.33 -> 6.77 | Improved |
| use a fine-tuned VLM backbone instead of a CLIP-I image encoder (ablation) | relsim · Relational retrieval | relational similarity score: 5.91 -> 6.77 | Improved |
| use a VLM instead of tuned pure vision encoders (CLIP, DINO) trained on same data (ablation) | relsim · Relational retrieval | relational similarity score | Improved |
| fine-tune CLIP and DINO on anonymous captions (ablation) | CLIP, DINO · Relational retrieval | relational similarity score | Improved |
| use DINO instead of LPIPS as similarity metric | image similarity metrics · Relational retrieval | relational similarity score: 4.56 -> 5.14 | Improved |
| use proprietary instead of open-source image editors | analogical generation · Analogical image generation | relsim: 0.71 -> 0.84 | Improved |
| use proprietary instead of open-source image editors | analogical generation · Analogical image generation | CLIP: 0.87 -> 0.78 | Hurt |
Closest papers by meaning
- Interpretable Measures of Conceptual Similarity by Complexity-Constrained Descriptive Auto-Encoding (CVPR 2024 · similarity 0.88)
- The Many Senses of Visual Similarity: A Text-Prompted Image Perceptual Metric (NeurIPS 2026 · similarity 0.87)
- VisMin: Visual Minimal-Change Understanding (NeurIPS 2024 · similarity 0.86)
- Composing Object Relations and Attributes for Image-Text Matching (CVPR 2024 · similarity 0.84)
- Identification of Necessary Semantic Undertakers in the Causal View for Image-Text Matching (AAAI 2024 · similarity 0.84)
- Spotting the Unseen: Reciprocal Consensus Network Guided by Visual Archetypes (AAAI 2024 · similarity 0.83)
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.
