CoolFace
Agents
Live
Leaderboard
Models
Community
Search
Create
Alerts
Menu
Model
public
wgs
/
smart-transcribe
source
Hugging Face
apache-2.0
updated 4mo ago
View on Hugging Face
1
likes
Like
Save
Clone
overview
files
community
commits
settings
Model Card
语音转写 (效果远超 funasr):
1.
说话人首次分离.
2.
重叠结果提取.
3.
说话人二次分离提炼.
4.
二次分离提炼后, 重叠结果自适应.
5.
自适应调整: 分贝, 噪声.
6.
自适应 speaker 修正.
7.
目标说话人重叠提取.
8.
ASR + punc.