First of all, thank you for releasing the code for such an interesting paper.
In the paper, it is mentioned that "we initialize the 3D convolution weights in U-Net with a pretrained 2D face restoration network" in section 3.3. Could you please let me know where it is implemented in the code? I appreciate your response.
First of all, thank you for releasing the code for such an interesting paper.
In the paper, it is mentioned that "we initialize the 3D convolution weights in U-Net with a pretrained 2D face restoration network" in section 3.3. Could you please let me know where it is implemented in the code? I appreciate your response.