achukhrov-ffr-team

> @cgeller if you have exported the model as saved model, then you can also use something like the flowing code for inference: > (test images are labeled img0.jpg to...

> @achukhrov-ffr-team Inference works well for me on the images from the camera unchanged. I have tried CLAHE (Adaptive histogram equalization) but it did not help much in my case....

@Steve-2040 Yes, but in train stage images (which were represented as same arrays with values from 0 to 255) , were preprocessed, specifically, divided by 255 in order to have...

@gdy2021, have you solved your problem? I have the same issue