🎙️

ASR / Speech Processing

Full-stack speech recognition: Whisper, Volcengine ASR, Alibaba ASR, supporting multilingual transcription and feature extraction.

Core Capabilities

Whisper large-v3 fine-tuning & deployment

Multi-ASR engine comparison & fusion

Audio feature extraction (encoder hidden states)

VAD endpoint detection & audio segmentation

🚧 More content coming soon...