You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
My video player (https://github.com/ggarra13/mrv2) is one of the top players in the VFX, CGI and animation industries. I would like to add SynthId image and/or video detection support to it to help in classifying deep fakes or potentially harmful AI generated content, since I understand Veo2 now supports such watermarking. Also, I am hoping other AI vendors follow suit in adding similar watermarking.
Can you point me in the right direction regarding code/implementation/paper? Or, better yet, a contact person at Google to collaborate with.
The text was updated successfully, but these errors were encountered:
Thanks for the note at LSB steganography. I am looking to detect AI videos, not create ones, so I am looking for something that might get used commercially, like SynthId.
LSB is too simple and very easy to bypass.
Thanks for the clarification. You might be able to get in touch with Google Cloud for more information about using SynthID to detect watermarks in videos. I've heard that more advanced SynthID software for detecting AI-generated content is given to select Vertex AI customers (e.g., for images). Note that this may not be the case for videos.
My video player (https://github.com/ggarra13/mrv2) is one of the top players in the VFX, CGI and animation industries. I would like to add SynthId image and/or video detection support to it to help in classifying deep fakes or potentially harmful AI generated content, since I understand Veo2 now supports such watermarking. Also, I am hoping other AI vendors follow suit in adding similar watermarking.
Can you point me in the right direction regarding code/implementation/paper? Or, better yet, a contact person at Google to collaborate with.
The text was updated successfully, but these errors were encountered: