Hello, 

I definitely agree: Automatic transcription would be an absolutely valuable tool as an addition to the subtitling possibilites within Trados. 

There are a lot of service providers active on the market that could be interesting as additional Apps for the AppStore, like of course, the speech-to-text  (STT) engines by Microsoft, Google and Amazon. But also very specialised powerful and speaker-independent automatic speech recognition (ASR) providers with high accuracy rates and options like custom dictionaries, speaker diarization and automatic punctuation - features that are of upmost importance using ASR engines.

Speechmatics, AppTek, Amazon offer solutions like that, only to mention a few of them. These are the most promising ones, I am following so far and which are also increasingly integrated and used by big subtitling software providers. I am not sure, how well Dragon NaturallySpeaking is able to perform besides the speaker-dependent application, which is really good for this use case and I work with Dragon on a daily basis. However, for generic, speaker-independent engines for audio-visual material, there would be some experience and tests interesting in comparison to the ones mentioned above.

I would be fantastic to see the integration different of ASR or STT engines via the AppStore for the subtitling app in Trados in the near future, as it is the case for the huge variety of MT engines you offer already! Therefore, I am really excited about your research regarding this topic. 

Regards

Hello, 

I definitely agree: Automatic transcription would be an absolutely valuable tool as an addition to the subtitling possibilites within Trados. 

There are a lot of service providers active on the market that could be interesting as additional Apps for the AppStore, like of course, the speech-to-text  (STT) engines by Microsoft, Google and Amazon. But also very specialised powerful and speaker-independent automatic speech recognition (ASR) providers with high accuracy rates and options like custom dictionaries, speaker diarization and automatic punctuation - features that are of upmost importance using ASR engines.

Speechmatics, AppTek, Amazon offer solutions like that, only to mention a few of them. These are the most promising ones, I am following so far and which are also increasingly integrated and used by big subtitling software providers. I am not sure, how well Dragon NaturallySpeaking is able to perform besides the speaker-dependent application, which is really good for this use case and I work with Dragon on a daily basis. However, for generic, speaker-independent engines for audio-visual material, there would be some experience and tests interesting in comparison to the ones mentioned above.

I would be fantastic to see the integration different of ASR or STT engines via the AppStore for the subtitling app in Trados in the near future, as it is the case for the huge variety of MT engines you offer already! Therefore, I am really excited about your research regarding this topic. 

Regards