Explicit 3D Representation
Explicit 3D representation aims to create detailed, readily manipulable 3D models, overcoming limitations of implicit methods like neural radiance fields (NeRFs) which are computationally expensive and difficult to integrate into existing 3D environments. Current research focuses on developing efficient algorithms, such as Gaussian splatting, and hybrid approaches combining implicit and explicit representations (e.g., using meshes to guide neural fields), to improve rendering quality, editing capabilities, and training speed. These advancements are significant for applications ranging from virtual and augmented reality to robotics and computer-aided design, enabling more realistic and interactive 3D experiences.
Papers
November 27, 2024
October 18, 2024
July 5, 2024
April 16, 2024
March 4, 2024
December 7, 2023
December 4, 2023
June 26, 2023
April 6, 2023
March 21, 2023
October 15, 2022