The world we live in consists of various 3D objects with their own shape characteristics. With advances in computational power, it becomes natural to understand our world in a 3D manner rather than 2D projection that limits full encoding of 3D objects. In this context, 3D shape analysis aims at the automatic analysis of 3D objects that can be naturally represented by 3D geometry. This field of research intersects subfields of computer science across computer vision, computer graphics, machine learning, and computational geometry. The applications in 3D shape analysis cover multidisciplinary areas such as medical imaging, biology, ergonomics, robotics, material/chemical engineering, etc. The main research projects include- Shape correspondence: development of surface registration techniques to establish a shape correspondence across complicated shapes (e.g., partially missing or highly variable structures) for population shape analysis- Shape annotation: development of automatic surface annotation techniques based on deep neural nets with new architecture design and data synthesis- Shape feature: development of automatic landmark extraction techniques to define unique/meaningful features (e.g., ridges and valleys) from surfaces for their structural variability analysis- Shape quantification: development of new shape quantification techniques to offer quantitative shape markersOur research team seeks generic shape analysis techniques. In its application domain, our team is in active collaboration with multiple PIs at Vanderbilt University Medical Center, UC-Berkeley Helen Wills Neuroscience Institute, and UNC-Chapel Hill Psychiatry for structural brain image research. The ultimate goal of this collaboration is to support understanding of human cognition/behaviors or brain diseases/disorders (Alzheimer disease, autism spectrum disorder, etc.).