With FaceBuilder, you don't need to become an experienced 3D modeler to create high-quality 3D models with clean topology. You first take several photos with neutral facial expressions from different angles, and then place a model on each photo to construct a head or facial model. The generated model can be used for rigid tracking using GeoTracker or for facial gesture tracking using FaceTracker. You can also export the model from Nuke for further modification in any other 3D software.
Even if there is only one photo, it can work
The more photos you have, the better the quality. But sometimes all you have is a frame in a photo or lens, and you still need to complete your work. Therefore, we have provided you with guarantees - you will be surprised that only one image can attract you
Three built-in topologies
Whether you need a high polygon model to shape as many facial features as possible, or a low polygon model to integrate it into some real-time workflows, we can meet your needs! FaceBuilder has low polygon, medium polygon, and high polygon topologies, allowing you to switch instantly!
Focal length estimation
Although we recommend manually setting camera settings to achieve maximum accuracy, we know that this is not always feasible. For this situation, you can use the "single focal length" and "variable focal length" estimation modes. When you know that the reference photo was taken with a camera and the same focal length, the first one is effective. If you are unsure whether it is a single camera or a zoom lens, the second one can provide assistance.
Automatic facial alignment driven by artificial intelligence
You no longer need to manually fix your face! At least not from scratch. Our facial alignment is not yet 100% accurate, so sometimes manual adjustments are required, but it fundamentally reduces the time required to shape the model. So much, we actually stopped manually executing this operation in the team during testing, and we don't want to go back!
Facial expression support
If you pursue ultimate quality and precision - you need photos with neutral facial expressions, but if you cannot obtain them, don't worry - FaceBuilder now supports FaceTracker's facial expressions, so even if you use non neutral photos, you can still achieve good results!
Complete diversity
Our growing statistical models already have data from the widest age range of humans from most parts of the world, so modeling people from anywhere in the world should not encounter any problems!
Configurable models
You can open and close different parts of the 3D model and customize them as needed. The shape is not affected, so if you decide to switch back when the model is ready - it won't damage anything!
Mixed Image Format
You can use photos with different aspect ratios in the reference sequence. If you know the camera settings for each photo in the sequence, you can set keyframes in the connected camera nodes. If you don't know the camera settings, simply turn on "Change Focal Length" in FaceBuilder settings, and the focal length of each photo in the sequence will be estimated.
