Research map / Detection and segmentation / CNN features
Accurate Scene Text Detection through Border Semantics Awareness and Bootstrapping
Chuhui Xue, Shijian Lu, Fangneng Zhan
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 |
|---|---|---|---|
| add semantics-aware text borders (ablation) | Baseline (ResNet) · MSRA-TD500 | f-score: 71.8 -> 74.3 | Improved |
| add bootstrapping augmentation (ablation) | Baseline (ResNet) · MSRA-TD500 | f-score: 71.8 -> 74.3 | Improved |
| add both borders and augmentation (ablation) | Border (ResNet) · MSRA-TD500 | f-score: 74.3 -> 76.8 | Improved |
| replace ResNet-50 with DenseNet-121 (ablation) | Aug.+Border · MSRA-TD500 | f-score: 76.8 -> 80.1 | Improved |
| add semantics-aware text borders (ablation) | Baseline (ResNet) · ICDAR2013 | f-score: 83.0 -> 84.9 | Improved |
| add bootstrapping augmentation (ablation) | Baseline (ResNet) · ICDAR2013 | f-score: 83.0 -> 85.2 | Improved |
| add both borders and augmentation (ablation) | Baseline (ResNet) · ICDAR2013 | f-score: 83.0 -> 87.4 | Improved |
| replace ResNet-50 with DenseNet-121 (ablation) | Aug.+Border · ICDAR2013 | f-score: 87.4 -> 89.2 | Improved |
| add semantics-aware text borders (ablation) | Baseline (ResNet) · ICDAR2017-RCTW | f-score: 58.5 -> 65.4 | Improved |
| add semantics-aware text borders (ablation) | Baseline (ResNet) · ICDAR2017-MLT | f-score: 62.6 -> 66.6 | Improved |
| add semantics-aware text borders and augmentation, evaluated at stricter IoU 0.8 vs 0.5 (ablation) | Baseline · MSRA-TD500 | f-score gap vs baseline: 5.0 -> 15.9 | Improved |
| add bootstrapping augmentation (faster convergence) (ablation) | Baseline (ResNet) · MSRA-TD500 | training time (hours): 56 -> 36 | Improved |
Closest papers by meaning
- Canny Text Detector: Fast and Robust Scene Text Localization Algorithm (CVPR 2016 · similarity 0.88)
- Multi-Oriented Scene Text Detection via Corner Localization and Region Segmentation (CVPR 2018 · similarity 0.87)
- Contextual Text Block Detection towards Scene Text Understanding (ECCV 2022 · similarity 0.87)
- Arbitrary Reading Order Scene Text Spotter with Local Semantics Guidance (AAAI 2025 · similarity 0.86)
- LayoutFormer: Hierarchical Text Detection Towards Scene Text Understanding (CVPR 2024 · similarity 0.86)
- LRANet: Towards Accurate and Efficient Scene Text Detection with Low-Rank Approximation Network (AAAI 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.
