Research map / Vision-language and multimodal / Human and embodied modeling
Fine-Grained Image-Text Matching by Cross-Modal Hard Aligning Network
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 hard assignment coding with a visual codebook instead of cross-attention (ablation) | CHAN, BiGRU · MS-COCO 5K test | RSUM: 418.6 -> 433.4 | Helped |
| use text queries and a textual codebook instead of a visual codebook (ablation) | CHAN, BiGRU · MS-COCO 5K test | RSUM: 433.4 -> 398.0 | Hurt |
| use LSE-Pooling rather than Max-Pooling to aggregate word-image similarities (ablation) | CHAN, BiGRU · MS-COCO 5K test | RSUM: 311.7 -> 433.4 | Helped |
| use LSE-Pooling rather than Average-Pooling (ablation) | CHAN, BiGRU · MS-COCO 5K test | RSUM: 431.9 -> 433.4 | Helped |
| use LSE-Pooling rather than Sum-Pooling (ablation) | CHAN, BiGRU · MS-COCO 5K test | RSUM: 428.1 -> 433.4 | Helped |
| use LSE-Pooling rather than Softmax-Pooling (ablation) | CHAN, BiGRU · MS-COCO 5K test | RSUM: 420.5 -> 433.4 | Helped |
| increase the visual codebook size (ablation) | CHAN, BiGRU · MS-COCO 5K test | RSUM | Helped |
| use CHAN rather than TERAN | BUTD + BERT · MS-COCO 5K test | RSUM: 443.4 -> 443.9 | Helped |
| use CHAN rather than VSE∞ | BUTD + BERT · Flickr30K test | RSUM: 513.5 -> 518.5 | Helped |
Closest papers by meaning
- Cross-Modal and Uni-Modal Soft-Label Alignment for Image-Text Retrieval (AAAI 2024 · similarity 0.89)
- Linguistic-Aware Patch Slimming Framework for Fine-grained Cross-Modal Alignment (CVPR 2024 · similarity 0.88)
- Identification of Necessary Semantic Undertakers in the Causal View for Image-Text Matching (AAAI 2024 · similarity 0.88)
- Text Embedding is Not All You Need: Attention Control for Text-to-Image Semantic Alignment with Text Self-Attention Maps (CVPR 2025 · similarity 0.87)
- Composing Object Relations and Attributes for Image-Text Matching (CVPR 2024 · similarity 0.87)
- Visual-Text Cross Alignment: Refining the Similarity Score in Vision-Language Models (ICML 2024 · similarity 0.87)
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.