Reproduce various text recognition model for OCR proposed in papers with TensorFlow implementation
relative with text recognition task
- [2018-ECCV] Synthetically Supervised Feature Learning for Scene Text Recognition
paper(keywords: SSFL) - [2018-CVPR] AON: Towards Arbitrarily-Oriented Text Recognition
paper(keywords: AON) - [2017-CVPR] Focusing Attention: Towards Accurate Text Recognition in Natural Images
paper(keywords: FAN) - [2017-NIPS] Gated Recurrent Convolution Neural Network for OCR
paper(keywords: GRCNN) - [2017-PAMI] An End-to-End Trainable Neural Network for Image-based Sequence Recognition and Its Application to Scene Text Recognition
paper(keywords: CRNN) - [2016-CVPR] Robust Scene Text Recognition with Automatic Rectification
paper(keywords: RARE) - [2016-IEEE] Recursive Recurrent nets with Attention Modeling for ocr in the wild
paper(keywords: R2AM)
| IIIT5k_50 | IIIT5k_1K | IIIT5k_None | SVT_50 | SVT_None | IC03_50 | IC03_full | IC03_50K | IC03_None | IC13_857_None | IC13_1015_None | IC15_None | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| CRNN | 97.6 | 94.4 | 78.2 | 96.4 | 80.8 | 98.7 | 97.6 | 95.5 | 89.4 | 86.7 (from SSFL) | 86.7 | - |
| SSFL | 97.3 | 96.1 | 89.4 | 96.8 | 87.1 | 98.1 | 97.5 | - | 94.7 | 94.0 | - | - |
| FAN | 99.3 | 97.5 | 87.4 | 97.1 | 85.9 | 99.2 | 97.3 | - | 94.2 | 93.9 (from SSFL) | 93.3 | 70.6 |
| GRCNN | 98.0 | 95.6 | 80.8 | 96.3 | 81.5 | 98.8 | 97.8 | - | 91.2 | - | - | - |
| RARE | 96.2 | 93.8 | 81.9 | 95.5 | 81.9 | 98.3 | 96.2 | - | 90.1 | 88.6 / 87.5 | - | - |
| R2AM | 96.8 | 94.4 | 78.4 | 96.3 | 80.7 | 97.9 | 97.0 | - | 88.7 | 90.0 ( from SSFL) | 90.0 | - |
command example :
python main.py --gpu 0 --model_name CRNN --testset ic13_857,ic13_1015,ic03_867 --optimizer RMSProp --loss ctc
python main.py --gpu 1 --model_name CRNN --testset ic13_857,ic13_1015,ic03_867 --optimizer Adam --loss ctc