AILS Talk: Annotation techniques for large-scale Volume EM datasets скачать в хорошем качестве

AILS Talk: Annotation techniques for large-scale Volume EM datasets

Our guest speaker Norma Rzepka was talking about annotation techniques. For many Machine Learning-based image analysis methods, the availability of high-quality ground truth data is still a necessity. In this talk, I will present some popular annotation techniques for volumetric electron microscopy (EM) datasets to obtain training or evaluation data. In the first part, the talk will cover different methods and tools to make dense volume annotations more efficient. The second part will be about collaborative workflows including quality control. The third part will cover sparse annotation apporaches that offer higher throughput than dense methods. Norman Rzepka is the co-founder of scalable minds. His company develops webKnossos, a collaborative web-based annotation tool for large-scale 3D image datasets, and services for automated segmentation of datasets. Norman has a background in computer science and is located in Potsdam, Germany.