zhangyuanhan-ai / visual_prompt_retrieval Goto Github PK

The SupPR models for each pascal-5i split is uploaded in link.But the last.pth file corresponding to folder2 and folder3 in this link is invalid and gives an error after downloading, is there a problem here?

Hello!

I am quite fascinated by your work and have gained significant knowledge from it. However, I am encountering some difficulty with the detection task. It appears that the support pairs have been preloaded. After reviewing the code you've published, I was unable to locate them.

Could you kindly assist me in generating the support set for detection task, similar to the one in the segmentation part?

SupPR.md mentions the command sh srun_train_pretrain.sh, but I can't find the srun_train_pretrain.sh file, is there a specific training file for that?

Hello, I am very interested in this job. But I am a little confused about the implementation of the detection part. When will the detection code be released?

Looking forward to your reply.

Hi, I met a bug in the running supervised prompt retrieval.
sh evaluate/srun_seg_evaluate.sh features_vit-laion2b_trn output_vit-laion2b-clip_trn

The step is followed with readme instructions. This bug is raised from loading the Rank50 prompt images calculated by python tools/calculate_similariity.py features_vit-laion2b trn trn. Interestingly, I smoothly run the unsupervised version and generated the segmentation results using the same code.

After comparing these two JSON file of top50-similarity in supervised and unsupervised versions, I notice their keys are different. Did I make mistakes?

does the input in line 10 - line 12 of tools/calculate_similariity.py mean sys.argv[1], sys.argv[2], sys.argv[3]?

the current code is:

eatures_name = sys.argv[1]
source_split = sys.argv[1]
target_split = sys.argv[2]

Thanks for your work, I had the same problem：
In the course of debugging get_positive_negative.py, I found that running the：
for img_name in iou_dict: if img_name in iou_dict[img_name]: del iou_dict[img_name][img_name]
Then iou_dict will become empty，like this：

So the sorted_iou and the sorted_iou_names is empty.
Are there any steps here that I'm missing?

Thank you for this inspiring work! I want to know which last.ckpt weight for VQGAN should be used in the Step3 of supervised prompt retrieval?

Hi, do you have any plan to release code for detection and colorization?

When I enter the python tools/featextrater_folderwise_UnsupPR.py vit_large_patch14_clip_224.laion2b features_vit-laion2b trn command, an error is reported:
huggingface_ hub.utils._errors.LocalEntryNotFoundError: An error happened while trying to locate the file on the Hub and we cannot find the requested files in the Local cache. Please check your connection and try again or make sure your Internet connection is on.
The server can't be networked to huggingface to download the pre-training weights, and there is no corresponding url link for the model.
How to solve this problem?