Learning a neural network requires serious computing power and time. We use our own cloud terminals, but calculating each video is still a very resource-intensive process. However, all the next videos with this very model will be calculated 10-15 times faster.
Once your video has been created, it is up to you to control the video. Your video has not been leaked from our side.
Yes, the model must be over 18 years old. The neural network is trained to identify content with minors. These videos will be deleted and the account blocked.
Progress cannot be stopped, it is as legitimate as any graphic or video editor can be. We provide a tool for creativity, and we are not responsible for the way it is used. You should understand that this is solely for your own use. It is your responsibility to do everything you do.
It's a complex legal issue. You create the video strictly for personal use. No public use. The law doesn't regulate the scope of fake videos yet, but you don't have to set a precedent. It's morally wrong to create these videos. But we're not here for morality, we're here to make a porn movie of our dreams.
Yes, we can. Faceshifter successfully works with two or more models. However, it requires much more power to calculate. And so we chose the optimal balance between performance, speed and quality. The POV format was chosen as the most suitable format for deepfake video fantasies.
Yes. That option is available. But it will be much more expensive, as it will require much more power. There will be a separate station dedicated to calculating your scene. The price will increase 5-7 times, as the neural network will be trained on a completely new video. There is also a limitation on the duration of the video, not exceeding 5 minutes.
This is a very extensive topic, which is difficult to implement within this site. We are currently developing a separate project dedicated to LGBT+.
FaceShifter is an advanced technology, we can work with even 1 single photo. But in machine learning technology, the calculation speed depends on the amount of raw materials. So the more photos the faster and better the final result. The most important thing is not the size of the photos, but the angles of the model's face. The more photos from different sides, the better DeepFake will be.