SenseVoice
SenseVoice is a locally running multilingual speech understanding tool. You can upload audio files to the app, which automatically completes speech recognition and outputs readable text based on the audio content. The application supports Chinese, English, Cantonese, Japanese, Korean, and also supports automatic language determination.
##Main functions - Multi-lingual speech recognition: Supports Chinese, English, Cantonese, Japanese, Korean and automatic recognition. - Audio upload recognition: Support selecting audio files from local devices or lazy cat network drives. - Batch upload: You can select multiple audio files at once, and the application will return the recognition results one by one. - Identification progress display: The actual upload progress is displayed during uploading, and the processing status and time consumption are displayed during the reasoning stage. - Rich text results: Keep the original markup and view information such as emotions, language and sound events. - API calls: Keep the `/docs` interface documents to facilitate integration through `/api/v1/asr`. The file selection portal has been connected to the lazy cat network disk automatic interception selector. When you click Select File, you can select: - Select from your local device: Upload audio files from your computer. - Select from the lazy cat online disk: directly select the audio files in the micro-service online disk. It is recommended to give priority to using audio within 30 seconds for testing. SenseVoice official Hugging Face example directory: huggingface.co/FunAudioLLM/SenseVoiceSmall/tree/main/example
