Shape Editing

Shape editing research focuses on developing methods for efficiently and intuitively modifying 3D shapes and their representations in images and videos. Current efforts concentrate on leveraging deep learning models, including diffusion models and transformers, to achieve precise localized edits guided by natural language, sketches, or explicit point cloud manipulation. These advancements are improving 3D content creation, facilitating more realistic and controllable avatar generation, and enhancing applications such as medical image analysis and surgical simulation. The resulting improvements in shape manipulation techniques are impacting various fields, from computer graphics and animation to medical procedures and product design.