Tencent Hunyuan releases Hy ASR 3.0 preview: speech recognition understands context, not just word-by-word
Tencent
Tencent Hunyuan has unveiled the preview of its Hy ASR 3.0 speech recognition model, which moves beyond literal transcription to grasp contextual meaning. The model is designed to improve accuracy and understanding in various real-world scenarios.
Tencent Hunyuan has released the preview version of its Hy ASR 3.0 speech recognition model. Unlike traditional systems that transcribe speech verbatim, Hy ASR 3.0 aims to understand the context and intent behind spoken words, leading to more accurate recognition. The model is part of Tencent's broader AI portfolio, which includes other models like the Hunyuan large language model. Hy ASR 3.0 preview is expected to enhance voice-based applications and services.
- Abbreviations
- ASR = Automatic Speech Recognition — автоматическое распознавание речи
Source: Tencent Hunyuan (GNews) —
original