Neural Geometry

Neural geometry focuses on representing and manipulating geometric information using neural networks, aiming to bridge the gap between classical geometry processing and modern machine learning. Current research emphasizes developing neural representations for surfaces and volumes, employing architectures like neural implicit surfaces and neural radiance fields, and applying these to tasks such as shape analysis, rendering, and multi-task learning. This interdisciplinary field is significant for advancing both neuroscience (understanding neural coding and perception) and computer graphics (efficient and realistic 3D modeling and rendering), with applications ranging from medical image analysis to robotics.