SegFormer
SegFormer copied to clipboard
NVlabs
Official PyTorch implementation of SegFormer
# Patching CVE-2007-4559 Hi, we are security researchers from the Advanced Research Center at [Trellix](https://www.trellix.com). We have began a campaign to patch a widespread bug named CVE-2007-4559. CVE-2007-4559 is a...
My train loss has always decrease, but the val_mIoU decreased at the first time, then it began increase. 
I found it diffcult to understand this two file. #1 local_configs/_base_/models/segformer.py #2 local_configs/_base_/models/segformer.py I knew that #1 means the base of #2, but I find the code in file 2:...
When I run a build of the docker container with the next command ``` sudo docker build -t mmsegmentation docker/ --no-cache ``` I receive the next errors related to the...
When I used segformer to directly use the Ade20k dataset for training and testing, it was shown that there was a difference in size. I tried to test_ Add resize...
Geforce RTX 3060 must use torch11.0+, and this means I must use cuda11.3, and the minimum version of mmcv is 1.4.7 So I got this error: "AssertionError: MMCV==1.4.7 is used...
Hi, in class MixVisionTransformer, function 'forward_features' return a list, and the classification head is a Linear module, I'm confused about the details of how you deal with it.
https://github.com/NVlabs/SegFormer/blob/639caf64aad8241c670c3a6a820ab0a9bde02b77/mmseg/models/backbones/mix_transformer.py#L172 As far as i understand it, self.num_patches is never used, but in my opinion the math behind its calculation is wrong. self.H and self.W are calculated in the line...
Have u update the trained models? Now the checkpoint on the downloading page can not be used on mmsegmentation. But the checkpoint which i downloaded last month can be used.
If the input is `HxWx3` (RGB Image) the output is `(H/4 x W/4 x num_classes)`. But the ground truth mask is `(H,W,num_classes)`. Can you point out the code where this...
