Depth images training

[resha1417]

Hello @mohan1914 ,

Thanks for the great work. In the paper you have mentioned that the adaptnet network should be trained on each modality separately. To train RGBD modality, did you just used RGBD images and corresponding labels or did you also used RGB Image information along with that .
What was the mean IoU when you trained with only RGBD modality.

Regards,
Resha Thacker