The article proposes a method named GauFRe for dynamic scene reconstruction using deformable 3D Gaussians, specifically designed for monocular video. The method extends the efficiency of Gaussian Splatting to accommodate…
Continue readingThe article introduces FLARE, a method for creating animatable and relightable 3D mesh avatars from a single monocular video. It aims to efficiently learn personalized 3D head avatars from videos…
Continue readingIn this research, the authors introduce a novel framework, MAMo, designed to enhance the accuracy of video depth estimation from monocular data. By incorporating memory and attention mechanisms, MAMo enables…
Continue readingLooks like you're using an ad blocker. Please support us by disabling these ad blocker.
