Sergey is a specialist in Machine Learning (Deep learning, Bayesian methods, text processing and much more) and Analysis of algorithms (Networking algorithms, Competitive analysis). An author of more than 100 scientific publications, several books and Author`s courses “Machine learning”, “Deep learning” and others. Also he has wide experience of cooperation with the industry in the field of Machine learning (Neuromation, SolidOpinion, Surfingbird, Deloitte Analytics Institute).
Topic: Deep CNN for the image segmentation
Short Description: Convolution Neural Networks have become the main modeling approach for images processing. We will discuss how the networks, which recognize individual objects, are growing into the networks allocating objects among the masses of others. We’ll start from the basics and talk about CNN in general, and than we will discuss architecture for segmentation: single-shot detectors and the line of models from R-CNN to Mask R-CNN.