search

jeremyxianx / RAWatermark

Official Implementation for: "RAW: A Robust and Agile Plug-and-Play Watermark Framework for AI-Generated Images (Videos) with Provable Guarantees"
MIT License
16 stars 1 forks source link

Some questions regarding the specific implementation #1

Open ZY123-GOOD opened 3 months ago

ZY123-GOOD commented 3 months ago

Hello, thank you for sharing this interesting work. I have a few questions. First, does this method require training for each image to obtain image-specific watermarks and classifiers? Second, do you have any plans to make the training part public to encourage researchers in the community to read and learn? I wish you all the best.

jeremyxianx commented 3 months ago

Hi, thank you for your interest in our work.

For your first question, if your goal is to train a watermark and its corresponding classifier specific to targeted images, the answer is yes. However, if you are referring to other use scenarios, please provide more details.

For the second question, yes, we are planning to release the full training code in the near future. Stay tuned (star the repo if you haven't).