Partial Shape Matching

Partial shape matching focuses on finding correspondences between incomplete 3D shapes, a crucial problem in computer vision and graphics with applications in areas like shape interpolation and object recognition. Current research emphasizes developing robust algorithms that address the challenges posed by missing data, often employing techniques like integer programming, functional maps, or neural networks to establish geometrically consistent correspondences between partial shapes. These methods aim to improve accuracy and efficiency compared to existing approaches, particularly in handling real-world scenarios with significant shape partiality, as demonstrated by superior performance on established and newly created benchmark datasets.