Research map / Detection and segmentation / CNN features
SwinTextSpotter: Scene Text Spotting via Better Synergy Between Text Detection and Text Recognition
Check 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 |
|---|---|---|---|
| add Recognition Conversion (ablation) | SwinTextSpotter · Total-Text | Detection H-mean: 82.8 -> 83.2 | Helped |
| add Recognition Conversion (ablation) | SwinTextSpotter · Total-Text | End-to-end H-mean: 63.4 -> 66.9 | Helped |
| add Recognition Conversion to the baseline (ablation) | ResNet-50 baseline · Total-Text | Detection H-mean: 78.9 -> 81.9 | Helped |
| add Recognition Conversion to the baseline (ablation) | ResNet-50 baseline · Total-Text | End-to-end H-mean: 55.7 -> 62.6 | Helped |
| replace ResNet-50 with Swin-Transformer (ablation) | SwinTextSpotter · Total-Text | End-to-end H-mean: 62.6 -> 65.5 | Helped |
| replace ResNet-50 with Swin-Transformer (ablation) | SwinTextSpotter · Total-Text | Detection H-mean: 81.9 -> 81.7 | No clear effect |
| add TLSAM (ablation) | SwinTextSpotter with Recognition Conversion · Total-Text | Detection H-mean: 81.9 -> 81.7 | No clear effect |
| add TLSAM (ablation) | SwinTextSpotter with Recognition Conversion · Total-Text | End-to-end H-mean: 62.6 -> 62.5 | No clear effect |
| add dilated convolution to Swin-Transformer (ablation) | SwinTextSpotter · Total-Text | Detection H-mean: 81.7 -> 82.4 | Helped |
| add dilated convolution to Swin-Transformer (ablation) | SwinTextSpotter · Total-Text | End-to-end H-mean: 65.5 -> 66.0 | Helped |
| add TLSAM and dilated convolution to the RC model (ablation) | SwinTextSpotter · Total-Text | Detection H-mean: 82.4 -> 83.2 | Helped |
| add TLSAM and dilated convolution to the RC model (ablation) | SwinTextSpotter · Total-Text | End-to-end H-mean: 66.0 -> 66.9 | Helped |
Closest papers by meaning
- SemiETS: Integrating Spatial and Content Consistencies for Semi-Supervised End-to-end Text Spotting (CVPR 2025 · similarity 0.89)
- Bridging the Gap Between End-to-End and Two-Step Text Spotting (CVPR 2024 · similarity 0.89)
- GoMatching: A Simple Baseline for Video Text Spotting via Long and Short Term Matching (NeurIPS 2024 · similarity 0.88)
- OTE: Exploring Accurate Scene Text Recognition Using One Token (CVPR 2024 · similarity 0.88)
- InstructOCR: Instruction Boosting Scene Text Spotting (AAAI 2025 · similarity 0.87)
- Arbitrary Reading Order Scene Text Spotter with Local Semantics Guidance (AAAI 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.