reCAPTCHA is a great system that allows the world to contribute to books' digitisation. I have tried a few times to build an OCR system. However, the lack of decent language models has always been a barrier. Does anyone know if it's possible to access the language models that have been built? |