Segmentation maps from ViTs

[DeepOceanDeep]

Hi @Muzammal-Naseer and @cgarbin and @kahnchana

I've been reading the paper and looking into the code too. But was not able to fins the piece of codes related to visualization of figures Like Figure 4: Attention maps and Figure 8: Segmentation maps from ViTs. Could you please give me some clue?

Thanks very much!

[cgarbin]

@DeepOceanDeep : I also a reader of the paper. I contributed a few PRs for the instructions, but have not worked on the original code.