Are there speech recognition and automatic translation programs for cuda and opencl?

Question

GPGPU

Hello! There is a large amount of speech processing ahead, I need to recognize English speech (the audio part on videos) and translate the resulting text into many languages. the software is good. but due to the large volume I would like to use gpu

Answer 1

The cheapest is to use cloud tools from Google or Yandex if you need the Russian language (they have the best technologies in this area)
In windows, using speech recognition, you can listen to English speech and translate it into text, translate it into another language with another tool, offline tools have (expensive).
All other options are so bad in quality that it makes sense to use them only from the point of view of the fan. True, I studied the issue somewhere about 5 years ago, maybe now it’s better.
ps if we are talking about illegal mechanisms, look for tools to automate actions on android, for example, the local google translator can work offline (it will almost certainly stop working after a certain amount of text, and you will have to somehow bypass it)
upd. tryhttps://github.com/mozilla/DeepSpeech