winner-4

Some questions about faceparsing

1

The current process of faceparsing is face detection + parsing. If I have a picture with a face cropped, can I use faceparsing directly without facer's face detection?

Incomplete segmentation results of Celebm model inference faceparsing

1

Is there any method to solve the problem of incomplete segmentation when inferring faceparsing results using the Celebm model provided by the author?

问下大家有没有现成的大量的闭眼人脸数据呢，有的话可以分享下吗，非常感谢！

有没有已经配好环境的大佬整个docker镜像呗，孩子已经被环境折磨了一天了