Frequently asked questions

General

What is VDR and is it Deepfakes?

Deepfake sarted as an open-source project for face-swapping and is now used as an umbrella term for video and audio manipulations using AI.

VDR stands for Video Dialogue Replacement and is not part of the Deepfakes project. With VDR you can change the dialgoue in a Target video, while keeping the lips and facial expressions in sync to a Source video.
With VDR you can replace the dialgoue in a video of Jimmy Fallon (Target) to be in sync with a dialgoue by John Oliver (Source).
With face-swapping you can make Jimmy Fallon's face look like John Oliver. The result is of the Target video looking similar to John Oliver, with the original diagoue of Jimmy Fallon.



Can I make anyone say anything?

No.
For each Target and Source video our team will first verify it complies with our Terms & Conditions, mainly to make sure the Target video is not of a famous persona and that the Source video does is not with malicious intent. For special use cases such as politicans for satirical content, or if you have the celebrity's approval please contact us.



Can I distribute the result wherever I want?

It depends on the details of the license you've purchased. The watermarked version is for personal review only, and the video may not be distributed anywhere online or presented to an external audience.



When do I need to decide if I want to purchase a license?

We want you to be satisfied with the result. Once you get the watermarked version, you will be able to choose which license you require and preferred payment method. Upon payment confirmation you will receive the non-watermarked version.



Can you assist me with finding a voice actor or selecting a Target video?

Sure! Please contact us if you require full service. For certain regions we will connect you with our local resellers that can provide you with a full service.



Is there a limit to how many watermarked videos I can create?

No, you can create how as many watermarked videos as you'd like and pay the minimal fee



Do you offer discounts for high volume usage?

Yes! If you plan on creating more than 30 minutes of videos per month please contact us directly.



I want to become a channel partner, can I do that?

Yes! Please contact us directly at partners@cannyai.com



Technical & Usage

Can I choose any Target video?

While we are always working to improve our technology, VDR is still an emerging technology. With such technologies, there are always limitations and restrictions so it would be possible to achieve optimal results. Once you will be given access to our platform, please make sure you read our guidelines.



What are Target and Source videos?

The Target video is a video of a person for which you would like to change the dialogue. This is the video our AI edits which you will see in the final result. The Source video is a video of the voice actor/actress, who is never shown in the final result. We only use his/her facial movements, lip movements and of course add the audio with the new dialgoue to the final result.



How long does it take?

After approval by our team of the Source and Target videos, for each pair the initial training and creation of the result video takes about 1-2 business days. If you are using a Target again but with a different Source dialgoue it will take up to 12 hours.



Is there a limit to the length of the Target or Source videos?

Yes, our system limits the Target and Source videos to be up to 5 minutes long. Please note that if the Source video is longer than the Target video the result will be limited by the Target's length.



Can I upload a video of more than one person?

While our system does support this option, it will not work directly through our platform. Please contact us if you require Full Service.



Does your AI work on Animation?

If the animation is lip-synced with good accuracy it should work. Please contact us for more information.