https://m.facebook.com/groups/107107546348803?view=permalink&id=413934485666106
As a lot of you read, Baidu has released their paper on speech-to-text called DeepSpeech. As written in paper, their end-to-end architecture offers 7x speed-up over previous architectures. And as I understand - sets the new state-of-the-art.
Papers are fun, but without data and code their hard to implement for a lot of individuals.
Great thing is that mozilla open-sourced DeepSpeech model that is implemented in Tensorflow. So now all of us can use it, tweak it, train it.
Gihub: https://github.com/mozilla/DeepSpeech
Blog: https://svail.github.io/mandarin/
Paper: http://jmlr.org/proceedings/papers/v48/amodei16.pdf
As a lot of you read, Baidu has released their paper on speech-to-text called DeepSpeech. As written in paper, their end-to-end architecture offers 7x speed-up over previous architectures. And as I understand - sets the new state-of-the-art.
Papers are fun, but without data and code their hard to implement for a lot of individuals.
Great thing is that mozilla open-sourced DeepSpeech model that is implemented in Tensorflow. So now all of us can use it, tweak it, train it.
Gihub: https://github.com/mozilla/DeepSpeech
Blog: https://svail.github.io/mandarin/
Paper: http://jmlr.org/proceedings/papers/v48/amodei16.pdf
'Deep Learning > resources' 카테고리의 다른 글
| RUBEDO: How to build and run your first deep learning network (0) | 2017.04.14 |
|---|---|
| 10 Free Must-Read Books for Machine Learning and Data Science (0) | 2017.04.12 |
| 10 minutes Practical TensorFlow Tutorial for quick learners (0) | 2017.04.08 |
| How Deep Neural Networks Work 동영상 강좌 24:37 (0) | 2017.04.05 |
| Object detection에 사용되는 RCNN, Fast RCNN, Faster RCNN 등을 간결하게 설명해놓은 깃북페이지 (0) | 2017.04.03 |
