mirrors/facefusion
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Use cv2.blur over cv2.GaussianBlur for better performance

Browse Source
This commit is contained in:
henryruhs 2023-10-30 22:45:10 +01:00
parent 857bede91e
commit bab666d6f9
1 changed files with 2 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
facefusion/face_helper.py
View File

@ -44,9 +44,9 @@ def paste_back(temp_frame: Frame, crop_frame: Frame, affine_matrix: Matrix) -> F
inverse_mask_size = int(numpy.sqrt(numpy.sum(inverse_crop_frame == 255)))
kernel_size = max(inverse_mask_size // 10, 10)
inverse_crop_frame = cv2.erode(inverse_crop_frame, numpy.ones((kernel_size, kernel_size)))
kernel_size = max(inverse_mask_size // 20, 5)
kernel_size = max(inverse_mask_size // 30, 5)
blur_size = kernel_size * 2 + 1
inverse_blur_frame = cv2.GaussianBlur(inverse_crop_frame, (blur_size , blur_size), 0) / 255
inverse_blur_frame = cv2.blur(inverse_crop_frame, (blur_size , blur_size)) / 255
inverse_blur_frame = numpy.reshape(inverse_blur_frame, [ temp_frame_height, temp_frame_width, 1 ])
temp_frame = inverse_blur_frame * inverse_temp_frame + (1 - inverse_blur_frame) * temp_frame
temp_frame = temp_frame.astype(numpy.uint8)