Getting plain white images instead of depth maps during inference

[YacineDeghaies]

the outputted depth value appear to be between 250 and 255. Is there a reason for this ?

Input Image:

ground truth depth map:

Inferred depth map: