High Quality Mesh
High-quality mesh generation and manipulation are central research areas in computer graphics and related fields, aiming to create accurate, efficient, and visually appealing 3D models. Current research focuses on developing novel neural network architectures, such as graph neural networks and diffusion models, to improve mesh reconstruction from various inputs (e.g., point clouds, images, implicit fields), optimize mesh properties (e.g., smoothness, watertightness), and enable efficient manipulation and animation. These advancements have significant implications for diverse applications, including virtual and augmented reality, computer-aided design, and scientific simulations, by providing more realistic and computationally tractable 3D representations.
Papers
MDNF: Multi-Diffusion-Nets for Neural Fields on Meshes
Avigail Cohen Rimon, Tal Shnitzer, Mirela Ben Chen
Volumetric Surfaces: Representing Fuzzy Geometries with Multiple Meshes
Stefano Esposito, Anpei Chen, Christian Reiser, Samuel Rota Bulò, Lorenzo Porzi, Katja Schwarz, Christian Richardt, Michael Zollhöfer, Peter Kontschieder, Andreas Geiger