2019/07/03 by Lukas Prantl, Prantl, Lukas, Nuttapong Chentanez +5 · 3 citations
Computer Science · Engineering · Mathematics · #3D Shape Modeling and Analysis #Advanced Vision and Imaging #Algorithm #Artificial intelligence #Artificial neural network #Astrophysics #Computer Graphics and Visualization Techniques #Computer Vision and Pattern Recognition (cs.CV) #Computer science #Dimension (graph theory) #FOS: Computer and information sciences #Feature (linguistics) #Flexibility (engineering) #Function (biology) #Geometry #Graphics (cs.GR) #Halo #Machine Learning (cs.LG) #Machine Learning (stat.ML) #Machine learning #Mathematics #Maxima and minima #Physics #Point (geometry) #Point cloud #Representation (politics) #Set (abstract data type) #Truncation (statistics) #cs.CV #cs.GR #cs.LG #stat.ML
paper · pdf · doi:10.48550/arxiv.1907.05279
published in arXiv (Cornell University) (Cornell University) · Further information and videos at https://ge.in.tum.de/publications/2020-iclr-prantl/
openalex publication_date 2019/07/03 · openalex created_date 2019/12/26 · arxiv created 2020/01/29 · arxiv updated 2020/01/30 · openalex updated_date 2026/08/05
Point clouds, as a form of Lagrangian representation, allow for powerful and flexible applications in a large number of computational disciplines. We propose a novel deep-learning method to learn stable and temporally coherent feature spaces for points clouds that change over time. We identify a set of inherent problems with these approaches: without knowledge of the time dimension, the inferred solutions can exhibit strong flickering, and easy solutions to suppress this flickering can result in undesirable local minima that manifest themselves as halo structures. We propose a novel temporal loss function that takes into account higher time derivatives of the point positions, and encourages mingling, i.e., to prevent the aforementioned halos. We combine these techniques in a super-resolution method with a truncation approach to flexibly adapt the size of the generated positions. We show that our method works for large, deforming point sets from different sources to demonstrate the flexibility of our approach.