Computer Vision News - July 2021

Workshop 24 Best of CVPR 2021 radiology, who have rich experience of AI applied to their respective context, and are optimistic about AI facilitating their clinical workflow in future. Dorothy Tzu-Chen Yen , Professor at Department of Nuclear Medicine and the Center for Advanced Molecular Imaging and Translation at Chang Gung University and Chang Gung Memorial Hospital, gave a wonderful talk on values of AI-based medical images, and presenting her recent series of new works on intelligent radiology image analysis from a clinician point of view. Philip Wai-Yan Chiu , Professor at Department of Surgery at The Chinese University of Hong Kong, gave an exciting talk that shared fantastic insights on the role of AI in endoscopy and minimally invasive surgery, in particular AI for standardization of endoscopic examination, lesion detection and characterization to meet the important clinical unmet need, and AI for enhancing therapeutic endoscopy for decision guidance, quality assurance and automation in procedure for the future. Elliot Fishman , Professor at Radiology, Oncology, Surgery and Urology at Johns Hopkins Hospital, gave a fantastic talk which showcased the role of AI in the earlier detection of pancreatic cancer through deep learning techniques and the combination of DL with radiomics features. In the last two sessions, Paul F. Jaeger , German Cancer Research Center, give a fantastic talk presenting nnU -Net which has achieved many success stories as a powerful automated design of deep learning method for biomedical image segmentation and is published in Nature Methods . Shadi Albarqouni , Helmholtz Zentrum München, gave a wonderful talk on deep federated learning in healthcare, with two recent novel works of FedDis to tackle the long-tail data distributions in different hospitals, and FedPerl to make better use of unlabeled data in semi-supervised federated learning. Kaleem Siddiqi ,