leonmakise / prores Goto Github PK

Hi, you have done an excellent job, and it would be perfect if you could include highly relevant references.
The following two papers demonstrate the effectiveness of pre-training visual prompts in the few-shot scenario and are closely related.
[1] VPPT: Visual Pre-Trained Prompt Tuning Framework for Few-Shot Image Classification
[2] PVP: Pre-trained Visual Parameter-Efficient Tuning

Hello! I'm really excited about your work for universal image restoration. I can't wait to see and run the code. Will the code be released soon? Thank you so much for your very significant contributions! :)

推理时，prompt是需要人为指定/输入的吗？

fantastic work !when the code will be released？

Hi, I find your paper fascinating, but when reading the paper, I I came across a particular sentence that I'm having difficulty understanding.

In Introduction, "the inference using the universal model is uncontrollable or unpredictable, as it cannot
guarantee that the output will meet our expectations without any task-specific indicator."

Isn't the purpose of a unified model to restore degraded images to clean ones? Why is it referred to as uncontrollable?