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