Research map / 3D vision / 3D generation
SPIN: Structure-Preserving Inner Offset Network for Scene Text Recognition
Chengwei Zhang, Yunlu Xu, Zhanzhan Cheng, Shiliang Pu, Yi Niu, Fei Wu, Futai Zou
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 chromatic SPIN rectification to the recognizer | VGG-7 STR baseline · Average across seven benchmarks | accuracy: 77.5 -> 78.8 | Improved |
| Add AIN to SPN to form full SPIN (ablation) | SPIN w/o AIN · Average across seven benchmarks | accuracy: +0.4% | Improved |
| Add AIN to SPN to form full SPIN (ablation) | SPIN w/o AIN · CUTE80 | accuracy: 63.5 -> 64.6 | Improved |
| Add chromatic SPIN before TPS-based STN | STN · Average across seven benchmarks | accuracy: +0.8% | Improved |
| Unify chromatic SPIN and TPS geometric rectification in GA-SPIN | SPIN+STN pipeline · Average across seven benchmarks | accuracy: +0.3% | Improved |
| Add GA-SPIN instead of geometric STN alone | VGG-7 STR baseline · Average across seven benchmarks | accuracy: +1.1% | Improved |
| Increase the number of intensity basis subspaces K (ablation) | SPIN and GA-SPIN · Regular and irregular benchmarks | mean accuracy | Mixed |
Later papers that compared with this method
34 later accepted papers report SPIN as a baseline in their results tables; 30 of them report a better number on most of the results they share with it.
- MobileOcc: A Human-Aware Semantic Occupancy Dataset for Mobile Robots (ECCV 2026 · better on 1 of 1 shared result)
- HybrIK-X: Hybrid Analytical-Neural Inverse Kinematics for Whole-Body Mesh Recovery (TPAMI 2025 · better on 2 of 2 shared results)
- Self-Supervised 3D Human Mesh Recovery from a Single Image with Uncertainty-Aware Learning (AAAI 2024 · better on 0 of 2 shared results)
- Playing for 3D Human Recovery (TPAMI 2024 · better on 2 of 2 shared results)
- On the Calibration of Human Pose Estimation (ICML 2024 · better on 1 of 1 shared result)
- PoseExaminer: Automated Testing of Out-of-Distribution Robustness in Human Pose and Shape Estimation (CVPR 2023 · better on 3 of 3 shared results)
- 3D Human Pose Estimation via Intuitive Physics (CVPR 2023 · better on 1 of 1 shared result)
- Multiview Human Body Reconstruction from Uncalibrated Cameras (NeurIPS 2022 · better on 3 of 4 shared results)
26 more not listed.
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
- ASTER: An Attentional Scene Text Recognizer with Flexible Rectification (TPAMI 2018 · similarity 0.87)
- Symmetry-Constrained Rectification Network for Scene Text Recognition (ICCV 2019 · similarity 0.87)
- Char-Net: A Character-Aware Neural Network for Distorted Scene Text Recognition (AAAI 2018 · similarity 0.86)
- Robust Scene Text Recognition With Automatic Rectification (CVPR 2016 · similarity 0.86)
- ESIR: End-To-End Scene Text Recognition via Iterative Image Rectification (CVPR 2019 · similarity 0.86)
- ContourNet: Taking a Further Step Toward Accurate Arbitrary-Shaped Scene Text Detection (CVPR 2020 · 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.
