CTC for Offline Handwriting Recognition
Автор: Oliver Nina
Загружено: 2016-02-02
Просмотров: 15510
Описание:
Overview of CTC algorithm for handwriting recognition as part of a paper presentation for the Family History Technology Workshop at Brigham Young University 2016. http://fhtw.byu.edu/
****UPDATE 02/10/18: Check our new open source code https://github.com/olivernina/nephi
Due to the number of inquiries about our method and code. We decided to release a new version of our code based on Pytorch and CRNN which link is above.
In recent handwriting recognition at ICFHR and ICDAR, CRNN has proven to be superior than a simpler feature selection described in this video, although the overall framework is still similar. A BYU team led by Dr. Barrett won the 2017 competition and their CRNN approach is currently the state of the art in this problem: https://www.computer.org/csdl/proceed...
We would like to thank the BYU Vision Lab under Dr. Bill Barrett and his students Curtis Wigington, Seth Stewart, etc for good discussions and collaboration.
Also thanks to fellow OSU student Russell Ault for his involvement and contribution to our project and Dr. Doug Kennard for his feedback and encouragement.
Повторяем попытку...
Доступные форматы для скачивания:
Скачать видео
-
Информация по загрузке:
![Эффект Джанибекова [Veritasium]](https://imager.clipsaver.ru/N9HlQ-XVnFk/max.jpg)
