Zero-shot-RIS
Zero-shot-RIS copied to clipboard
[CVPR 2023] Official code for "Zero-shot Referring Image Segmentation with Global-Local Context Features"
Is your method tied to a particular CLIP, or can it be easily switched to other CLIP variants, e.g. EVA-CLIP or other CLIPs in OpenCLIP?
Dear Authors, Thank you for your valuable research and open code. I was trying to reproduce your repository following the instructions. However, at the third step of installation, the download...
I would like to ask what type of GPU your team used when doing experiments, how many Gpus were used, and how long the training time was.
Got the following error when performing `pip install -e` on the `old-detectron2` directory. ``` ~/miniconda3/envs/zsref/lib/python3.8/site-packages/torch/utils/cpp_extension.py:782: UserWarning: The detected CUDA version (11.5) has a minor version mismatch with the version that...
While running the notebook KCCV2023_tutorial.ipynb (I have created the environment), I am getting errors due to module not found , I tried to change the module name according to folder...
Dear author, I would like to ask you how to implement the qualitative analysis of the effectiveness of your global-local context visual features in the paper? Specifically , if we...
Thank you for this great work. When checking the code, I found that in file "Our_method_with_free_solo.py", 'sentence_raw' (line 71) often has more than 1 diverse sentence while 'target' is just...
Thank you for such an inspiring work! Could you please provide the Our_method_with_free_solo for the Phrasecut dataset? Thanks!