Open caiexu opened 5 years ago
Could you send source code which used to be for generating landmark feature image of the white figure to me?
The generated face is very blurred based on my landmark feature image when I test. However, the generated face image so clear using your test data downloaded from GitHub.
As for the following show: the file named aa.jpg is drawn by my code and image named 1.jpg is generated based on aa.jpg. And others is your data.
Looking forward to your reply Best regards
Thank you so much
Hi, Caiexu
The landmark could be generated by using MTCNN and TCDCN methods as mentioned in the paper.
thank you for replying Maybe I can not describe the question clearly. I mean could you send the code to me? the cold is not the program that generates the landmark feature the cold is the program that creates the image with black dots on a white background in which dots represent the landmark feature. the code is the program that creates training dataB could you send the program to me?
thank you for replying Maybe I can not describe the question clearly. I mean could you send the code to me? the cold is not the program that generates the landmark feature the cold is the program that creates the image with black dots on a white background in which dots represent the landmark feature. the code is the program that creates training dataB could you send the program to me?
Hi, Caiexu I could not found the exactly script I written, but the link from TCDCN should be helpful. https://github.com/zhzhanp/TCDCN-face-alignment
Could you send source code which used to be for generating landmark feature image of the white figure to me?
The generated face is very blurred based on my landmark feature image when I test. However, the generated face image so clear using your test data downloaded from GitHub.
As for the following show: the file named aa.jpg is drawn by my code and image named 1.jpg is generated based on aa.jpg. And others is your data.
Looking forward to your reply Best regards
Thank you so much