-
Duo character consistency in video
I am looking for a video generation tool/platform/model/whatever does je job, that is capable of keeping 2 consistent characters based on input photo. I want a video with 2 people (me and my girlfriend) to make a Christmas tree. With nano banana I already have the starting photo (living room -> add Christmas tree in progress + add the 2 people from the other photos keeping their appearances). So next step is to get it to video but so far the faces are changing way too much to resemble me and my girlfriend anymore…..the Christmas tree making is perfect, just the faces are a problem. I think that this is cutting edge stuff that I cannot do with comfyui anymore at this moment and I am willing to spend some credits at a commercial platform for this as well.
Any tips would be wildly appreciated!
replied 5 months, 3 weeks ago 2 Members · 1 Reply
-
1 Reply
-
Hey Hans,
Sounds exciting.
You are already suggesting trying other platforms.
The A.I. Generative market is moving so fast now, that one provider that might be the best today will be the worst tomorrow. Which is where the likes of AVID are now giving access to multiple platforms for the user to choose from (subscribe) for a specific job. Whether generative or utility A.I..
My advice is to think about workflow, in the same way you’ll use compositing tools to bring multiple sources together in one video/image.
Currently I would:
Consider Elevenlabs.io for voice cloning, if you want your voice on it?
For video, the developers with the deepest pockets would be:
https://www.sora-2.studio/
and
https://veo3ai.io/But there may be an outlier which is better. Or you may find different tools does different parts better.
Hope that this helps?
Atb
Mads
sora-2.studio
Sora 2 Studio – AI Video Generator with Native Audio & Lip Sync
Create cinematic videos with OpenAI's Sora 2. Generate realistic videos with native audio, perfectly synced lip movements, and lifelike physics from text or images.
Reply to this Discussion! Login or Sign Up