Paint-it: Text-to-texture synthesis via deep convolutional texture map optimization and physically-based rendering
We present Paint-it a text-driven high-fidelity texture map synthesis method for 3D meshes
via neural re-parameterized texture optimization. Paint-it synthesizes texture maps from a …
via neural re-parameterized texture optimization. Paint-it synthesizes texture maps from a …
Mononphm: Dynamic head reconstruction from monocular videos
Abstract We present Monocular Neural Parametric Head Models (MonoNPHM) for dynamic
3D head reconstructions from monocular RGB videos. To this end we propose a latent …
3D head reconstructions from monocular RGB videos. To this end we propose a latent …
Imitator: Personalized speech-driven 3d facial animation
Speech-driven 3D facial animation has been widely explored, with applications in gaming,
character animation, virtual reality, and telepresence systems. State-of-the-art methods …
character animation, virtual reality, and telepresence systems. State-of-the-art methods …