The main features of xzf-thu/mega-asr are: Speech Processing, Speech Recognition.
Open-source alternatives to xzf-thu/mega-asr include: openai/whisper — This project is a speech recognition and translation engine that utilizes a sequence-to-sequence transformer… stepfun-ai/step-audio2. facebookresearch/omnilingual-asr — Omnilingual-ASR is a multilingual automatic speech recognition framework and toolkit designed to transcribe audio… kyutai-labs/delayed-streams-modeling — Kyutai's Speech-To-Text and Text-To-Speech models based on the Delayed Streams Modeling framework. qwenlm/qwen3-asr. bytedance/megatts3 — MegaTTS3 is a bilingual speech synthesis system that generates natural-sounding speech in Chinese and English,…
Kyutai's Speech-To-Text and Text-To-Speech models based on the Delayed Streams Modeling framework.
This project is a speech recognition and translation engine that utilizes a sequence-to-sequence transformer architecture to convert audio into text. It is built upon a weakly supervised learning framework, which leverages large-scale, unlabelled audio-transcript data to create generalized speech representations capable of performing simultaneous transcription, language identification, and translation. The system distinguishes itself through a unified multi-task modeling approach that shares token sequences across different objectives, allowing it to handle diverse languages and vocabularies
Omnilingual-ASR is a multilingual automatic speech recognition framework and toolkit designed to transcribe audio across 1,600 languages. It provides a complete pipeline for converting speech to text, including a toolkit for fine-tuning pre-trained speech models to specific languages or datasets using custom training recipes. The system supports zero-shot speech recognition, allowing the model to predict text in unseen languages without extensive training data. It further enables few-shot language guidance through in-context examples and uses language codes to constrain transcription output t
