Research map / Detection and segmentation / CNN features
Detecting Text in Natural Image with Connectionist Text Proposal Network
Zhi Tian, Weilin Huang, Tong He, Pan He, Yu Qiao
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 |
|---|---|---|---|
| refine RPN proposals with a Fast R-CNN model (Faster R-CNN) | RPN · ICDAR 2013 (component study) | F-measure: 0.27 -> 0.75 | Improved |
| predict fine-scale 16-pixel text proposals instead of whole text lines (FTPN) | Faster R-CNN · ICDAR 2013 (component study) | F-measure: 0.75 -> 0.80 | Improved |
| add in-network BLSTM recurrent connection (ablation) | FTPN (no RNN) · ICDAR 2013 (component study) | F-measure: 0.80 -> 0.88 | Improved |
| add in-network BLSTM recurrent connection (ablation) | FTPN (no RNN) · ICDAR 2013 (component study) | precision: 0.83 -> 0.93 | Improved |
| add in-network BLSTM recurrent connection (ablation) | FTPN (no RNN) · ICDAR 2013 (component study) | recall: 0.78 -> 0.83 | Improved |
| add in-network BLSTM recurrent connection (ablation) | CTPN · ICDAR 2013 (test) | running time per image (s): 0.13 -> 0.14 | No clear effect |
| add side-refinement (ablation) | CTPN · SWT and Multilingual | F-measure: about 2% improvement | Improved |
| reduce input short side from 600 to 450 | CTPN · ICDAR 2013 (test) | F-measure: 0.88 -> 0.84; 0.14s -> 0.09s | Mixed |
| CTPN versus prior best TextFlow and FASText | TextFlow · ICDAR 2013 (test) | F-measure: 0.80 -> 0.88 | Improved |
| CTPN versus prior best method | Zhang · ICDAR 2015 | F-measure: 0.54 -> 0.61 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- Faster R-CNN — Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks (TPAMI 2016) · better on 3 of 3 shared results
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
- Towards End-To-End Text Spotting With Convolutional Recurrent Neural Networks (ICCV 2017 · similarity 0.90)
- Multi-Oriented Text Detection With Fully Convolutional Networks (CVPR 2016 · similarity 0.89)
- EAST: An Efficient and Accurate Scene Text Detector (CVPR 2017 · similarity 0.88)
- An End-to-End TextSpotter With Explicit Alignment and Attention (CVPR 2018 · similarity 0.88)
- Single Shot Text Detector With Regional Attention (ICCV 2017 · similarity 0.88)
- CPN: Complementary Proposal Network for Unconstrained Text Detection (AAAI 2024 · 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.
