Issue with testing a tif image

[srikrvd]

Hi Issam,

I trained the LCFCN for a cell counting use case. When I test a TIF image, the model gives the following error,
RuntimeError: output with shape [1, 2048, 2048] doesn't match the broadcast shape [3, 2048, 2048]
But when I convert the same image to a JPG, the model predicts the count. On the downside, since I convert to a JPG, there is some information loss and the cell count is not accurate.
May I know why is this happening and how to solve this?

Thanks,
Srikanth

[IssamLaradji]

Is there a way you can make the TIF image have 3 channels in the first dimension without having to convert it to JPG?