We introduce the first neural optimization framework to solve a classical instance of the tiling problem. Namely, we seek a non-periodic tiling of an arbitrary 2D shape using one or more types of tiles: the tiles maximally fill the shape's interior without overlaps or holes. To start, we reformulate tiling as a graph problem by modeling candidate tile locations in the target shape as graph nodes and connectivity between tile locations as edges. Further, we build a graph convolutional neural network, coined TilinGNN, to progressively propagate and aggregate features over graph edges and predict tile placements. TilinGNN is trained by maximizing the tiling coverage on target shapes, while avoiding overlaps and holes between the tiles. Importantly, our network is self-supervised, as we articulate these criteria as loss terms defined on the network outputs, without the need of ground-truth tiling solutions. After training, the runtime of TilinGNN is roughly linear to the number of candidate tile locations, significantly outperforming traditional combinatorial search. We conducted various experiments on a variety of shapes to showcase the speed and versatility of TilinGNN. We also present comparisons to alternative methods and manual solutions, robustness analysis, and ablation studies to demonstrate the quality of our approach.
Weakly Supervised Spatio-Temporal Candidate Discovery of Dairy Farm Sites from Seasonal Satellite Imagery
School of Computer Science, University of Galway（戈尔韦大学计算机科学学院）；Faculty of Computer Science and Engineering, GIK Institute of Engineering Sciences and Technology（工程科学与技术GIK学院计算机科学与工程系）
DeepSMILE: Contrastive self-supervised pre-training benefits MSI and HRD classification directly from H&E whole-slide images in colorectal and breast cancer
Netherlands Cancer Institute（荷兰癌症研究所）；University of Amsterdam（阿姆斯特丹大学）；Ellogon AI B.V.（Ellogon人工智能有限公司）；Radboud University Medical Center（拉德堡德大学医学中心）
Mapping the landscape of histomorphological cancer phenotypes using self-supervised learning on unlabeled, unannotated pathology slides
University of Glasgow（格拉斯哥大学）；New York University School of Medicine（纽约大学医学院）；New York University Langone Health（纽约大学朗格尼医学中心）；Karolinska Institutet（卡罗林斯卡学院）；University College London Hospital（伦敦大学学院医院）
Magnification-Aware Distillation (MAD): A Self-Supervised Framework for Unified Representation Learning in Gigapixel Whole-Slide Images
Center for Applied Artificial Intelligence, University of Kentucky（应用人工智能中心，肯塔基大学）；Sanders-Brown Center on Aging, University of Kentucky（老龄化桑德斯-布朗中心，肯塔基大学）；Department of Electrical and Computer Engineering, University of Kentucky（电气与计算机工程系，肯塔基大学）；Division of Neuropathology, University of Pittsburgh（神经病理学部，匹兹堡大学）；Department of Biomedical Informatics, Emory University（生物医学信息学系，埃默里大学）；Department of Pathology and Laboratory Medicine, University of California Davis（病理学与实验室医学系，加州大学戴维斯分校）；Department of Laboratory Medicine and Pathology, University of Texas Health（实验室医学与病理学系，德克萨斯大学健康科学中心）
Using Whole Slide Image Representations from Self-Supervised Contrastive Learning for Melanoma Concordance Regression
Proscia, Inc.（Proscia公司）；University of Florida College of Medicine（佛罗里达大学医学院）；Sidney Kimmel Medical College at Thomas Jefferson University（托马斯·杰斐逊大学西德尼·金梅尔医学院）
