Research map / Vision-language and multimodal / Human and embodied modeling
Text Matching as Image Recognition
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 |
|---|---|---|---|
| use hierarchical convolution over a word-level matching matrix | MatchPyramid · MSRP | accuracy: 69.90 -> 75.94 | Helped |
| use hierarchical convolution over a word-level matching matrix | MatchPyramid · Paper citation matching | accuracy: 86.48 -> 88.73 | Helped |
| use an indicator matching matrix instead of dot-product similarities (ablation) | MatchPyramid · MSRP | accuracy: 75.94 -> 75.77 | Hurt |
| use cosine similarity instead of dot-product similarities (ablation) | MatchPyramid · MSRP | accuracy: 75.94 -> 75.13 | Hurt |
| use an indicator matching matrix instead of dot-product similarities (ablation) | MatchPyramid · Paper citation matching | F1: 82.86 -> 44.71 | Hurt |
| use cosine similarity instead of dot-product similarities (ablation) | MatchPyramid · Paper citation matching | accuracy: 88.73 -> 86.65 | Hurt |
| include word-vector norms in the matching similarity (ablation) | MatchPyramid · Paper citation matching | accuracy: 86.65 -> 88.73 | Helped |
Closest papers by meaning
- Expressing Objects Just Like Words: Recurrent Visual Embedding for Image-Text Matching (AAAI 2020 · similarity 0.88)
- An End-to-End TextSpotter With Explicit Alignment and Attention (CVPR 2018 · similarity 0.87)
- Similarity Reasoning and Filtration for Image-Text Matching (AAAI 2021 · similarity 0.86)
- Composing Object Relations and Attributes for Image-Text Matching (CVPR 2024 · similarity 0.84)
- Expanding the Scope of Negatives: Boosting Image-Text Matching with Negatives Distribution Guided Learning (AAAI 2025 · similarity 0.84)
- CoMat: Aligning Text-to-Image Diffusion Model with Image-to-Text Concept Matching (NeurIPS 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.
