| --- |
| license: apache-2.0 |
| language: |
| - zh |
| - en |
| pipeline_tag: audio-classification |
| tags: |
| - speaker-verification |
| - speaker-diarization |
| - FunASR |
| - CAM++ |
| library_name: funasr |
| --- |
| |
| # CAM++ |
|
|
| **Speaker Verification & Diarization** — identify and distinguish speakers in audio. |
|
|
| CAM++ is a speaker embedding model used for speaker verification (is this the same person?) and speaker diarization (who spoke when?). |
|
|
| ## Quick Start |
|
|
| ```python |
| from funasr import AutoModel |
| |
| # Speaker diarization as part of ASR pipeline |
| model = AutoModel( |
| model="funasr/paraformer-zh", |
| hub="hf", |
| vad_model="funasr/fsmn-vad", |
| punc_model="funasr/ct-punc", |
| spk_model="funasr/campplus", |
| device="cuda", |
| ) |
| result = model.generate(input="meeting.wav") |
| # Each sentence has a speaker label |
| for sentence in result[0]["sentence_info"]: |
| print(f"[Speaker {sentence['spk']}] {sentence['text']}") |
| ``` |
|
|
| ## Features |
|
|
| - Speaker embedding extraction |
| - Speaker verification (same/different speaker) |
| - Speaker diarization (multi-speaker segmentation) |
| - Works with FunASR pipeline via `spk_model` parameter |
|
|
| ## Model Details |
|
|
| | Property | Value | |
| |----------|-------| |
| | Architecture | CAM++ (Class-Aware Multi-scale) | |
| | Embedding Dim | 192 | |
| | Sample Rate | 16kHz | |
|
|
| ## Links |
|
|
| - **GitHub**: [FunASR](https://github.com/modelscope/FunASR) |
| - **Docs**: [modelscope.github.io/FunASR](https://modelscope.github.io/FunASR/) |
|
|
