Joint Normal

Joint normal processing, encompassing tasks like depth map super-resolution and point cloud denoising, aims to improve the quality and resolution of 3D data by leveraging relationships between points and their associated surface normals. Current research emphasizes the use of deep learning models, particularly convolutional neural networks and transformers, often incorporating joint filtering techniques or contrastive learning frameworks to simultaneously refine both point positions and normal vectors. This integrated approach enhances accuracy and detail preservation compared to separate processing, with applications impacting various fields including computer vision, 3D modeling, and robotics.