4 ms·
I checked and it's using Mozilla's DeepSpeech library: https://github.com/MozillaReality/FirefoxReality/blob/dee6f420265825b9e0849e8ab35ffbc52eac88e2/app/src/co
by tedmielczarek 8y ago
I checked and it's using Mozilla's DeepSpeech library:
https://github.com/MozillaReality/FirefoxReality/blob/dee6f420265825b9e0849e8ab35ffbc52eac88e2/app/src/common/shared/org/mozilla/vrbrowser/ui/VoiceSearchWidget.java#L22 https://github.com/MozillaReality/FirefoxReality/blob/dee6f4...
which is here: https://github.com/mozilla/DeepSpeech https://github.com/mozilla/DeepSpeech
DeepSpeech is the actual voice recognition library, CommonVoice is a project to improve its accuracy by gathering more voice data.
- skykooler 8y agoInteresting. The DeepSpeech github indicates that it's not capable of realtime transcription, even using a GPU. How is that fast enough for this use case?