Thank you for this bypass project. Using the latest v4-round6, after completing all deployments and applying 5 images, although the built-in watermark detector in this project gave good results for all of them, when I actually uploaded output.png to Gemini and used the command "@Synthid" to test it, the result still said that all of them were "generated or edited by Google AI". I am sure that I used all the steps and tools correctly, and I tried both final and Nuke, but neither worked. I wonder if the author has used your tool to upload to Gemini for official testing?
Thank you for this bypass project. Using the latest v4-round6, after completing all deployments and applying 5 images, although the built-in watermark detector in this project gave good results for all of them, when I actually uploaded output.png to Gemini and used the command "@Synthid" to test it, the result still said that all of them were "generated or edited by Google AI". I am sure that I used all the steps and tools correctly, and I tried both final and Nuke, but neither worked. I wonder if the author has used your tool to upload to Gemini for official testing?