twke18 / hsg Goto Github PK

Hi, great work!

However, I am little confused about your comparison against SOTA methods. For instance, wrt cityscapes, why don't you compare against STEGO? Is it because you are computing scores on cityscapes val set while STEGO on test set? If yes, would you then compare against this method: https://paperswithcode.com/sota/unsupervised-semantic-segmentation-on-1? Also, do you have results on cityscapes test set?

Thanks.

I used a GPU card on the server to train with your open source code, and the result was stuck in this training progress. I strictly configured the environment you gave, using pytorch==1.12, python==3.7.9, may I ask you What are your thoughts on this issue?

Hi author, the following error occurs when executing the train.sh file for the cityscapes dataset:

The gtFineInstId file is mentioned in the datasets/cityscapes/val.txt file (shown below), but it is not present in the Cityscapes dataset constructed according to your request, how can this be resolved?

Hello! Thanks for your job.

I would like to know how to train the original Segsort to compare with your method. Because the original source code of Segsort was basing on Tensorflow, did you reproduced the Segsort on Pytorch for comparsion?

Hi author! Thanks for sharing this great work!

I have a question about how to find the same semantic group across view? Can you help me by pointing out where the corresponding code is?

Thanks a lot!

Hi author, I have downloaded the Cityscapes Trained Models and the file is as follows:

I would like to know how to set up the file to perform inference on the Cityspaces dataset with the model you provided.

thanks.