Research map / Detection and segmentation / CNN features
SPTS v2: Single-Point Scene Text Spotting
Yuliang Liu, Jiaxin Zhang, Dezhi Peng, Mingxin Huang, Xinyu Wang, Jingqun Tang, Can Huang, Dahua Lin, Chunhua Shen, Xiang Bai, Lianwen Jin
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 |
|---|---|---|---|
| share parameters between IAD and PRD (ablation) | SPTS v2 · Total-Text Full lexicon | end-to-end accuracy: 80.0 -> 81.4 | Improved |
| integrate token and feature in information transmission (ablation) | SPTS v2 · Total-Text Full lexicon | end-to-end accuracy: 78.4 -> 81.4 | Improved |
| add location feature to information transmission (token alone vs both) (ablation) | SPTS v2 · Total-Text lexicon-free | end-to-end accuracy: 66.4 -> 68.5 | Improved |
| use central point instead of top-left corner (ablation) | SPTS · Total-Text lexicon-free | end-to-end accuracy: 71.6 -> 74.2 | Improved |
| use central point instead of random point (ablation) | SPTS · Total-Text lexicon-free | end-to-end accuracy: 73.2 -> 74.2 | Improved |
| single point instead of Bezier box (ablation) | SPTS v2 · Total-Text lexicon-free | end-to-end accuracy: 63.2 -> 75.5 | Improved |
| single point instead of rectangle box (ablation) | SPTS v2 · SCUT-CTW1500 lexicon-free | end-to-end accuracy: 55.0 -> 63.6 | Improved |
| double training schedule for Bezier representation (ablation) | SPTS-Bezier · SCUT-CTW1500 lexicon-free | end-to-end accuracy: 52.6 -> 51.1 | Hurt |
| random instance order instead of area order (ablation) | SPTS · SCUT-CTW1500 Full lexicon | end-to-end accuracy: 75.3 -> 83.8 | Improved |
| deeper decoder (6 layers instead of 1) (ablation) | SPTS v2 R50 · Total-Text lexicon-free | end-to-end accuracy: 53.7 -> 68.5 | Improved |
| larger backbone ResNet-34 instead of ResNet-18 (ablation) | SPTS v2 · Total-Text lexicon-free | end-to-end accuracy: 60.6 -> 64.7 | Improved |
| larger backbone ResNet-50 instead of ResNet-34 (ablation) | SPTS v2 · Total-Text lexicon-free | end-to-end accuracy: 64.7 -> 68.5 | Improved |
Later papers that compared with this method
2 later accepted papers report SPTS v2 as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.
- Arbitrary Reading Order Scene Text Spotter with Local Semantics Guidance (AAAI 2025 · better on 2 of 2 shared results)
- SwinTextSpotter v2: Towards Better Synergy for Scene Text Spotting (IJCV 2025 · better on 0 of 1 shared result)
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 3 of 3 on most shared results.
- SwinTextSpotter — SwinTextSpotter: Scene Text Spotting via Better Synergy Between Text Detection and Text Recognition (CVPR 2022) · better on 2 of 2 shared results
- MANGO — MANGO: A Mask Attention Guided One-Stage Scene Text Spotter (AAAI 2021) · better on 1 of 1 shared result
- ABCNet v2 — ABCNet: Real-Time Scene Text Spotting With Adaptive Bezier-Curve Network (CVPR 2020) · better on 1 of 1 shared result
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- DeepSolo: Let Transformer Decoder With Explicit Points Solo for Text Spotting (CVPR 2023 · similarity 0.91)
- SemiETS: Integrating Spatial and Content Consistencies for Semi-Supervised End-to-end Text Spotting (CVPR 2025 · similarity 0.90)
- OTE: Exploring Accurate Scene Text Recognition Using One Token (CVPR 2024 · similarity 0.89)
- Bridging the Gap Between End-to-End and Two-Step Text Spotting (CVPR 2024 · similarity 0.88)
- GoMatching: A Simple Baseline for Video Text Spotting via Long and Short Term Matching (NeurIPS 2024 · similarity 0.88)
- 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.
