Hi,YiF.I have read the literature with reference to your open source code in the past two days, and have the following questions about the Region Encoder part:

1. H^W=N, then N is the number of tokens, do h and w change with the changes of H and W?
2. For a 512^512 image, H^h=512, W^w=512?
3. h and w are both 4 in your source code, so for a 512^512 image, H=W=128 and N=16384, is that too much?

亲爱的作者你们好，非常感谢你们的贡献！ 我在运行 test.py 时遇到了如下的错误：

File "/home/disk/xxx/git/RegionProxy/models/vit.py", line 113, in forward x, hw_shape = self.patch_embed(inputs), (self.patch_embed.DH, File "/home/xxx/anaconda3/envs/region_proxy/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1185, in getattr raise AttributeError("'{}' object has no attribute '{}'".format( AttributeError: 'PatchEmbed' object has no attribute 'DH'

我运行的参数为：

--config configs/regproxy_cityscapes/regproxy-s16-sub4+implicit-mid-2+768x768+80k+adamw-poly+cityscapes.py --checkpoint checkpoints/regproxy-s16-sub4+implicit-mid-2+768x768+80k+adamw-poly+cityscapes.pth --eval mIoU

以下为我的文件目录结构：

├── checkpoints │   └── regproxy-s16-sub4+implicit-mid-2+768x768+80k+adamw-poly+cityscapes.pth ├── configs │   ├── base │   │   ├── datasets │   │   ├── default_runtime.py │   │   ├── models │   │   └── schedules │   ├── regproxy_ade20k │   │   ├── regproxy-b16-sub4+implicit-mid-4+512x512+160k+adamw-cr+ade20k.py │   │   ├── regproxy-l16-sub4+implicit-mid-9+640x640+160k+adamw-cr+ade20k.py │   │   ├── regproxy-r26-s32-sub4+implicit-mid-n1+512x512+160k+adamw-poly+ade20k.py │   │   ├── regproxy-r50-l32-sub4+implicit-mid-n1+640x640+160k+adamw-cr+ade20k.py │   │   ├── regproxy-s16-sub4+implicit-mid-4+512x512+160k+adamw-poly+ade20k.py │   │   └── regproxy-t16-sub4+implicit-mid-4+512x512+160k+adamw-poly+ade20k.py │   └── regproxy_cityscapes │   ├── regproxy-b16-sub4+implicit-mid-2+768x768+80k+adamw-cr+cityscapes.py │   ├── regproxy-l16-sub4+implicit-mid-5+768x768+80k+adamw-cr+cityscapes.py │   ├── regproxy-s16-sub4+implicit-mid-2+768x768+80k+adamw-poly+cityscapes.py │   └── regproxy-t16-sub4+implicit-mid-2+768x768+80k+adamw-poly+cityscapes.py ├── data │   └── cityscapes │   ├── gtFine │   ├── leftImg8bit │   ├── test.txt │   ├── train.txt │   └── val.txt ├── LICENSE ├── models │   ├── init.py │   ├── proxy_head.py │   ├── segmentors.py │   └── vit.py ├── README.md ├── test.py ├── train.py └── utils ├── checkpoint.py ├── init.py

Hi, thanks for your code! There are some about handle borders in your code. And I'm not very clear on the purpose of this. Could you please explain that? Thank you very much!

1、 About h,w The sentence "(Hh) × (W w) matches the size of the output segmentation map and (h, w) is the relative stride of the initial token gird" in the paper indicate that h,w is the downsample stride of segmentation map, but when I reading the code, I feel confused how it works, throught rerange the token_logits and matrix multiplication we get the final segmentation map,which is as large as the input image. So why do you set the extra parameter h and w, and how do h,w relate with stride?

2、About F.unfold() Official Implement Code token_logits = F.unfold(token_logits, kernel_size=3, padding=1).reshape(B, -1, 9, H, W) # (B, C, 9, H, W) pseudocode in the paper # get neighbors for each cell y = rar(y, "B N K -> B K H W") nb = im2col(y, kernel_size=3, padding=1) nb = rar(nb, "B (K n) (H W) -> B H W n K") The other is what does F.unfold() do in the code ,in the paper ,you show the process of proxy head using pseudocode，and say im2col( i.e. F.unfold() ) is using to get neighbors for each cell, I can not understand this well ,too.

Looking forward to your reply!!! Thank you ~~~

Dear authors, Thanks for sharing your great work. I want to know how to visualize the regions as shown in Fig.6, and Fig 8. Can you release the code?

Thanks

Hi,YiF.Based on your open source code, I implemented an unofficial implementation of RegionProxy based only on Pytorch.This is the link of it[https://github.com/wanghr-git/RegionProxy](url).Can you check it for correctness if you have time?

Hi, YIF@YiF-Zhang If my dataset is 512*512 resolution, how can I use the pretrained model? Can the pre-trained model at 224 or 384 resolution be used directly?