niceshot

sorry to bother, I wonder if you could publich the code to compute MIOU and AP

Pretrained segmentation model on cityscapes?

3

sorry to bother, I wonder if you could please publish your pretrained segmentation model on cityscapes

So many packages are lost

6

I wonder if it can run

Could anyone tells me what this sentence means : attn_qs = attn[torch.arange(B)[:, None], patch_id, :]; why use "torch.arange(B)"

Questions about distill_loss

3

sorry to bother, I see the distill_loss in distill.py as : distill_loss = F.kl_div( F.log_softmax(distill_logits / T, dim=-1), F.softmax(teacher_logits / T, dim=-1).detach(), reduction='batchmean') I wonder why the teacher part uses...

question about MultiBoxLoss and the Model

1

To be honest, this is definitely the best tutorial in object detection I have ever read, thanks for your sharing ! The question I want to ask is that I...

I tried to print them but it seems the values are wrong, could anyone tell me the right value for each class?

questions about swd

6

Sorry to bother, I saw you used SWD (sliced wasserstein distance) in the paper, and I also want to use this score as a metric for some tests. Could you...

code running

1

Is there anyone who successfully runs this code?