RuntimeError: Given groups=1, weight of size [16, 3, 3, 3], expected input[1, 4, 345, 514] to have 3 channels, but got 4 channels instead

[suitmyself]

RuntimeError: Given groups=1, weight of size [16, 3, 3, 3], expected input[1, 4, 345, 514] to have 3 channels, but got 4 channels instead

[sjmoran]

Check your input image. It appears to have 4 channels rather than 3. The 4th is possibly an alpha channel. If that is the case just select the first 3 channels which should be the RGB channels.

[suitmyself]

I just run the inference with the image provided by this repo.
It seems that pillow.Image.open read image with 'RGBA' 4 channels, not 'RGB' 3 channels.
So I add 'convert('RGB')' as follows:
img = ImageProcessing.normalise_image(np.array(Image.open(img_filepath).convert('RGB')), normaliser)