Alibaba releases Fun-ASR 1.5, a large-scale speech recognition model.

CoinFeed reported on April 20 that Alibaba today released a new generation of end-to-end speech recognition model, achieving high-precision recognition of 30 languages ​​with a single model, covering the seven major dialect systems of Chinese and more than 20 local accents, and strengthening the special recognition of recitation of ancient poems.