okooooooo/EasyVtuber

0 Fork 0

T

yuyuyzl 973edfa4be readme

2022-04-24 18:20:39 +08:00

readme

2022-04-24 18:20:39 +08:00

UPDATE README.md / Add Sample Character Images

2021-12-16 01:52:40 +09:00

DELETE half precision

2022-04-07 23:40:19 +09:00

.gitignore

ADD Half precision TalkingHeadAnime

2021-12-22 18:25:17 +09:00

env_conda.yaml

env

2022-04-24 16:44:41 +08:00

facial_points.py

ADD facial_points.py / EDIT pose.py To change use facial_points.py

2021-12-13 02:08:22 +09:00

LICENSE

Create LICENSE

2021-12-19 14:26:18 +09:00

main.py

env

2022-04-24 16:44:41 +08:00

models.py

merge main.py

2021-12-10 17:42:06 +09:00

pose.py

ADD facial_points.py / EDIT pose.py To change use facial_points.py

2021-12-13 02:08:22 +09:00

README.md

readme

2022-04-24 18:20:39 +08:00

utils.py

ADD utils.py comment

2021-12-10 17:41:35 +09:00

README.md

EasyVtuber

Fork自 https://github.com/GunwooHan/EasyVtuber
为解决面捕质量问题，又反向port了原版demo https://github.com/pkhungurn/talking-head-anime-2-demo 中关于ifacialmocap的ios面捕逻辑
并且省略了ifacialmocap pc端，通过UDP直连的方式使ios面捕刷新率达到30fps，解决了面捕刷新率的瓶颈
最后，将EasyVtuber中使用的OBS虚拟摄像头方案切换为Unity Capture，解锁RGBA输出能力，无需绿背即可直接使用