by Sarah Frigg | Jan 3, 2023 | Capture, Digital Humans, Machine Learning
Self-Supervised Effective Resolution Estimation with Adversarial Augmentations We demonstrate that our method outperforms state-of-the-art image quality assessment methods in estimating the sharpness of real and generated human faces. January 3, 2023 IEEE Winter...
by Sarah Frigg | Nov 30, 2022 | Capture, Digital Humans, Machine Learning
Production-Ready Face Re-Aging for Visual Effects We demonstrate how the simple U-Net, surprisingly, allows us to advance the state of the art for re-aging real faces on video, with unprecedented temporal stability and preservation of facial identity across variable...
by Sarah Frigg | Oct 11, 2022 | Rendering, Video Processing, Visual Computing
TempFormer: Temporally Consistent Transformer for Video Denoising We propose an efficient hybrid Transformer-based model, TempFormer, which composes SpatioTemporal Transformer Blocks (STTB) and 3D convolutional layers. October 11, 2022European Conference on Computer...
by Sarah Frigg | Oct 5, 2022 | Capture, Digital Humans, Machine Learning
Learning Dynamic 3D Geometry and Texture for Video Face Swapping We approach the problem of face swapping from the perspective of learning simultaneous convolutional facial autoencoders for the source and target identities, using a shared encoder network with...
by Sarah Frigg | Sep 13, 2022 | Capture, Digital Humans, Machine Learning
Facial Animation with Disentangled Identity and Motion using Transformers We propose a 3D+time framework for modeling dynamic sequences of 3D facial shapes, representing realistic non-rigid motion during a performance. September 13, 2022ACM/Eurographics Symposium...
