viz3 = np.vstack((255*tgt_img_viz, 255*depth_viz, 255*mask_viz... in test_mask

[gagajian]

Hello, there seems something wrong with "vstack". The codes reports error as follows:

viz3_im = Image.fromarray(viz3.astype('uint8')) TypeError: Cannot handle this data type

Can you help me with this, thanks!

[JohnnieXDU]

you should change 'viz3.astype('uint8')' to 'viz3.transpose(1, 2, 0).astype('uint8')', due to the format should be [h, w, 3], not [3, h, w].

[lotus0903]

you should change 'viz3.astype('uint8')' to 'viz3.transpose(1, 2, 0).astype('uint8')', due to the format should be [h, w, 3], not [3, h, w].

Through this, I also got the issue about typeerror and the black flow images. Do you know the reason? Thanks a lot.

[anuragranj]

You need to make sure of the following. Maybe the conventions have changed with pytorch versions.

1. If you are using the correct dim order HWC or CHW.
2. If you in the correct range, there is a multiplication by 255 which assumes that the expected range should be 0-255. However if conventions have changed to 0-1 then avoid this multiplication.