FunASR Chinese Recognition:

FunASR is a set of open-source speech recognition models from Alibaba. It performs better than the Whisper series in Chinese speech scenarios. Video translation software already supports its use through HTTP calls via the zh_recogn and SenseVoice projects. You only need to deploy the corresponding zh_recogn and SenseVoice integration package, and after starting it, fill in the API address in the video translation settings.

However, many users are still confused about this operation. Therefore, starting from version v2.97, this function has been integrated into the video translation software. This means you no longer need to deploy and start the zh_recogn and SenseVoice projects separately. You can directly select FunASR Chinese Recognition in the software.

Select FunASR Chinese in Speech Recognition

After selecting FunASR Chinese Recognition in Speech Recognition, you can choose to use the paraformer-zh model or the SenseVoiceSmall model. It is recommended to choose the former, as it offers better performance and speed than the latter.

Download Model Online for the First Time Using FunASR Chinese Recognition

To avoid a large package size, the FunASR models are not integrated into the software package. The first time you use it, it will automatically download from modelscope.cn and save it to the hub folder under the models folder in the software directory. Depending on your network conditions, this may take from a few minutes to tens of minutes or even longer. As long as there are no red errors, be patient and wait for the download to complete.