Research map / Vision-language and multimodal / Human and embodied modeling
Knowledge Mining With Scene Text for Fine-Grained Recognition
Hao Wang, Junchao Liao, Tianheng Cheng, Zewen Gao, Hao Liu, Bo Ren, Xiang Bai, Wenyu Liu
Check 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 |
|---|---|---|---|
| Add KnowBert knowledge-enhanced text features to ViT | ViT image classifier · Con-Text | mAP: 79.24 -> 89.53 | Improved |
| Add KnowBert knowledge-enhanced text features to ViT | ViT image classifier · Drink Bottle | mAP: 80.81 -> 85.26 | Improved |
| Add KnowBert knowledge-enhanced text features to ViT | ViT image classifier · Crowd Activity | mAP: 80.29 -> 87.45 | Improved |
| Add KARC to the ViT + BERT baseline (ablation) | ViT + BERT · Con-Text | mAP: 86.51 -> 87.25 | Improved |
| Add KARC to the ViT + BERT baseline (ablation) | ViT + BERT · Drink Bottle | mAP: 82.81 -> 83.59 | Improved |
| Add KARC to the ViT + BERT baseline (ablation) | ViT + BERT · Crowd Activity | mAP: 85.34 -> 86.16 | Improved |
| Add VKAC on top of KARC (ablation) | Model A with KARC · Con-Text | mAP: 87.25 -> 89.53 | Improved |
| Add VKAC on top of KARC (ablation) | Model A with KARC · Drink Bottle | mAP: 83.59 -> 85.26 | Improved |
| Add VKAC on top of KARC (ablation) | Model A with KARC · Crowd Activity | mAP: 86.16 -> 87.45 | Improved |
| Train the full model end-to-end instead of separately (ablation) | ViT + KnowBert classifier · Con-Text | mAP: 81.47 -> 89.53 | Improved |
| Train the full model end-to-end instead of separately (ablation) | ViT + KnowBert classifier · Drink Bottle | mAP: 82.26 -> 85.26 | Improved |
| Train the full model end-to-end instead of separately (ablation) | ViT + KnowBert classifier · Crowd Activity | mAP: 81.79 -> 87.45 | Improved |
Closest papers by meaning
- Improving Scene Graph Classification by Exploiting Knowledge from Texts (AAAI 2022 · similarity 0.88)
- Democratizing Fine-grained Visual Recognition with Large Language Models (ICLR 2024 · similarity 0.87)
- FG-CLIP: Fine-Grained Visual and Textual Alignment (ICML 2025 · similarity 0.85)
- The Devil is in the Fine-Grained Details: Evaluating Open-Vocabulary Object Detectors for Fine-Grained Understanding (CVPR 2024 · similarity 0.85)
- GUIDED: Granular Understanding via Identification, Detection, and Discrimination for Fine-Grained Open-Vocabulary Object Detection (NeurIPS 2025 · similarity 0.85)
- OST: Refining Text Knowledge with Optimal Spatio-Temporal Descriptor for General Video Recognition (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.
