DIS
DIS copied to clipboard
xuebinqin
This is the repo for our new project Highly Accurate Dichotomous Image Segmentation
Hello author, I have been using the A100 to train the ISNet model. However, I encountered a problem with CUDA incompatibility. To resolve this issue, I spent some time upgrading...
Hello, there is some issue with Downloading the weigh files from Google Drive, and the other one requires some password or account. Can you please make the Google Drive link...
First of all, thank you very much for your work. When ’interm_sup’ is set to True, the encoder will be trained. My data set has tens of thousands of images....
Hello, thanks for the great work and nice code. I am wondering how to provide the ground truth data in case of a 2-class segmentation problem. I used a segmentation...
Hello, @xuebinqin, thanks for open sourcing your work. Could you please share some training tips and answer below questions? I am using default configuration: ``` hypar["early_stop"] = 20 hypar["batch_size_train"] =...
Hi, Thanks for this great work and for uploading the trained weights. I'm using your pretrained model for inference and I don't really get the purpose of the **cache_size** parameter....
Thanks for your great work. what are the training differences between isnet.pth and isnet-general-use.pth, as isnet-general-use works much better than isnet.pth.
Thanks for your great work! What is the final value of training loss that can be achieved by using intermediate supervision loss？
Hello, I hope someone is able to help me here as I am completely stuck. I get the follwoing error when running the inference: ValueError: Can't write images with one...
