- Task 1 - Video mode Localisation
- Task 2 - Video Mode End-to-End
- Task 3 - Still image mode Localisation
- Task 4 - Still image mode Word Recognition
- Task 5 - Still image mode End-to-End
Inactive evaluations
method: Clova AI / Lens2018-10-05
Authors: Seolki Baek, Geonmo Gu, Jeongo Seo
Description: Description: Our model is featured by CNN/RNN-based encoder and Hybrid CTC/Attention decoder. Moreover we proposed new text synthesis tools to make our model robust and high performance in the wild.
method: CLOVA-AI2018-11-30
Authors: Junyeop Lee, Jeonghun Baek, Sungrae Park, Moonbin Yim, Seonghyeon Kim, Hwalsuk Lee
Description: Attention based model
method: DictNet2017-09-02
Authors: RRC DOST organizers
Description: Synthetic Data and Artificial Neural Networks for Natural Scene Text Recognition
M Jaderberg, K Simonyan, A Vedaldi, A Zisserman
NIPS Deep Learning Workshop 2014
Description Paper Source Code
| Date | Method | Total Edit distance (case sensitive) | Correctly Recognised Words (case sensitive) | T.E.D. (case insensitive) | C.R.W. (case insensitive) | |||
|---|---|---|---|---|---|---|---|---|
| 2018-10-05 | Clova AI / Lens | 12.8571 | 90.30% | 11.6667 | 90.91% | |||
| 2018-11-30 | CLOVA-AI | 37.0734 | 66.67% | 32.7083 | 69.09% | |||
| 2017-09-02 | DictNet | 126.5288 | 13.33% | 103.2199 | 24.24% | |||
| 2017-07-24 | google vision api | 122.3921 | 18.79% | 121.8063 | 18.79% | |||
| 2017-08-31 | tesseract 4.00 (LSTM) | 138.4040 | 18.18% | 132.3580 | 18.79% |