Research map / Vision-language and multimodal / Cross-modal learning
Saliency-Guided Attention Network for Image-Sentence Matching
Zhong Ji, Haoran Wang, Jungong Han, Yanwei Pang
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 |
|---|---|---|---|
| replace global visual feature with SVA saliency feature (ablation) | SAN (no STA) · MSCOCO 1K test | sentence retrieval R@1: 63.4 -> 65.4 | Improved |
| replace global visual feature with SVA saliency feature (ablation) | SAN (no STA) · MSCOCO 1K test | image retrieval R@1: 50.5 -> 52.9 | Improved |
| fuse global visual feature with SVA feature (ablation) | SAN (no STA) · MSCOCO 1K test | sentence retrieval R@1: 65.4 -> 66.2 | Improved |
| fuse global visual feature with SVA feature (ablation) | SAN (no STA) · MSCOCO 1K test | image retrieval R@1: 52.9 -> 53.7 | Improved |
| use self-attention textual feature (IT) on top of fused visual feature (ablation) | SAN (FV + GT) · MSCOCO 1K test | image retrieval R@1: 53.7 -> 56.6 | Improved |
| merge global and self-attention textual features (FT G-I) (ablation) | SAN (FV + GT) · MSCOCO 1K test | sentence retrieval R@1: 66.2 -> 66.8 | No clear effect |
| add STA module (FT G-S) to fused visual baseline (ablation) | SAN (FV + GT) · MSCOCO 1K test | sentence retrieval R@1: 66.2 -> 85.4 | Improved |
| add STA module (FT G-S) to fused visual baseline (ablation) | SAN (FV + GT) · MSCOCO 1K test | image retrieval R@1: 53.7 -> 69.1 | Improved |
| use SVA feature instead of global feature as STA visual guidance (ablation) | SAN (STA) · MSCOCO 1K test | sentence retrieval R@1: 74.5 -> 82.1 | Improved |
| use SVA feature instead of global feature as STA visual guidance (ablation) | SAN (STA) · MSCOCO 1K test | image retrieval R@1: 57.8 -> 67.3 | Improved |
| fuse global and SVA visual features as STA guidance (ablation) | SAN (STA) · MSCOCO 1K test | sentence retrieval R@1: 82.1 -> 85.4 | Improved |
| fuse global and SVA visual features as STA guidance (ablation) | SAN (STA) · MSCOCO 1K test | image retrieval R@1: 67.3 -> 69.1 | Improved |
Closest papers by meaning
- Instance-Aware Image and Sentence Matching With Selective Multimodal LSTM (CVPR 2017 · similarity 0.89)
- Dual Attention Networks for Multimodal Reasoning and Matching (CVPR 2017 · similarity 0.89)
- Multi-Modality Cross Attention Network for Image and Sentence Matching (CVPR 2020 · similarity 0.88)
- Top-Down Visual Saliency Guided by Captions (CVPR 2017 · similarity 0.88)
- Temporal Saliency Query Network for Efficient Video Recognition (ECCV 2022 · similarity 0.88)
- Language-Guided Salient Object Ranking (CVPR 2025 · 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.
