As a VFX Producer working on a film or television series, I find the current verification workflow completely impractical for professional production use.
When a production has secured written consent from an actor granting the right to use their likeness for the purposes of a specific project, that should be sufficient. Are you telling me that written legal authorization is not enough — and that I need to physically bring the actor to my facility just to pass a verification check?
And what happens when I need a different angle of the same actor? Do I call them in again? What about when I need to run multiple tests using different reference photos of the same person — different ages, with or without makeup, with glasses, wearing a hat — do I schedule a new verification session every single time?
This makes absolutely no sense in a professional filmmaking context. The current system seems designed for social media influencers making fun videos, not for legitimate cinematic and broadcast production pipelines where legal clearances are already in place and industry workflows must be respected.
Furthermore, even if I were to overlook all of the above and ask the actor to go through your verification process, what happens if they refuse? And frankly, why wouldn't they? I would have to explain to them that their full facial biometric data will be stored in your company's database — data that is among the most sensitive and personal information a human being has. Why would any actor agree to have their face ID permanently held by a third-party company they have no relationship with and no control over?
I am not really sure if I got this right. So I want to use a photo e.g. of my grandma. Next the "Real human verification" is solved by showing my face via webcam to verify that I am a real human and then my identity will be stored and linked to the uploaded photo of my grandma?? And will the photo I want to use with Seedance (e.g. my grandma) be verified on its on (is the uploaded image a VIP / IP character / Is the user really authorized to use the photo)?
Clarification please <3 – I know it is supposed to avoid misuse … but at the same time, this whole process makes me feel pretty uncomfortable.
With all due respect, this must be the STUPIDEST implementation of Seedance2.0 on ANY platform.
Unless its only meant for playing around with text2video, which is 98% of the time useless to professional workflows. AI-generated actors simply do not work with Seedance2 in ComfyUI; using ANY of the Seedance2 workflows. Only option is to go to some web service instead. Way to go!
I created a virtual character in AI, and I'm still getting back the same error message: Exception: API Error: The request failed because the input image may contain real person. Request id: 02177776830765534cb6184a92ddd9ae87302c189a666e59fa55a (Type: BadRequest)
You really need to fix this! Especially since I also used Runway with the exact same image and it generated perfectly there. I prefer control and ComfyUI should be offering this as it does with other nodes.
As a VFX Producer working on a film or television series, I find the current verification workflow completely impractical for professional production use.
When a production has secured written consent from an actor granting the right to use their likeness for the purposes of a specific project, that should be sufficient. Are you telling me that written legal authorization is not enough — and that I need to physically bring the actor to my facility just to pass a verification check?
And what happens when I need a different angle of the same actor? Do I call them in again? What about when I need to run multiple tests using different reference photos of the same person — different ages, with or without makeup, with glasses, wearing a hat — do I schedule a new verification session every single time?
This makes absolutely no sense in a professional filmmaking context. The current system seems designed for social media influencers making fun videos, not for legitimate cinematic and broadcast production pipelines where legal clearances are already in place and industry workflows must be respected.
Furthermore, even if I were to overlook all of the above and ask the actor to go through your verification process, what happens if they refuse? And frankly, why wouldn't they? I would have to explain to them that their full facial biometric data will be stored in your company's database — data that is among the most sensitive and personal information a human being has. Why would any actor agree to have their face ID permanently held by a third-party company they have no relationship with and no control over?
I am not really sure if I got this right. So I want to use a photo e.g. of my grandma. Next the "Real human verification" is solved by showing my face via webcam to verify that I am a real human and then my identity will be stored and linked to the uploaded photo of my grandma?? And will the photo I want to use with Seedance (e.g. my grandma) be verified on its on (is the uploaded image a VIP / IP character / Is the user really authorized to use the photo)?
Clarification please <3 – I know it is supposed to avoid misuse … but at the same time, this whole process makes me feel pretty uncomfortable.
With all due respect, this must be the STUPIDEST implementation of Seedance2.0 on ANY platform.
Unless its only meant for playing around with text2video, which is 98% of the time useless to professional workflows. AI-generated actors simply do not work with Seedance2 in ComfyUI; using ANY of the Seedance2 workflows. Only option is to go to some web service instead. Way to go!
Maybe it's just me. The template in Comfy Cloud doesn't work. I'm getting an unsupported node pack error for the ByteDanceAsset nodes.
Super excited with this update
I created a virtual character in AI, and I'm still getting back the same error message: Exception: API Error: The request failed because the input image may contain real person. Request id: 02177776830765534cb6184a92ddd9ae87302c189a666e59fa55a (Type: BadRequest)
You really need to fix this! Especially since I also used Runway with the exact same image and it generated perfectly there. I prefer control and ComfyUI should be offering this as it does with other nodes.