jianfch / stable-ts

Transcription, forced alignment, and audio indexing with OpenAI's Whisper
MIT License
1.43k stars 165 forks source link

How to achieve real-time highlighted words in the demo video? #372

Open Karterlar opened 1 month ago

Karterlar commented 1 month ago

I'm evaluating the accuracy of the timestamp, which is important to me。

jianfch commented 1 month ago

The demo videos are not in real-time. They are the transcription results exported as subtitle files then encoded with their respective audios. See https://github.com/jianfch/stable-ts/issues/364#issuecomment-2143532786.