Courses
Computer vision
Course is dedicated to 2D computer vision methods. Program includes image formation process, image filtering, local features detection and matching, image retrieval, robust fitting methods, introduction to machine learning and deep learning, image categorization, object detection, image segmentation, optical flow estimation, background subtraction, object tracking, action recognition. Lectures are accompanied by seminars and practical tasks.The course is taught in Moscow State University and Yandex School of Data Analysis.
Media compression
This course takes part in autumn semester and is about different compression algorithms for various mediums such as text, photo, and video. Algorithms that are introduced in the course include both classical and modern machine learning approaches. Lectures are supplemented with tasks in C/C++ and Python so attendees can utilize gained knowledge in practice. Approximately top 5 students with the best performance are taken to the lab on a trial basis. More information in telegram channel @VGCourse
Video processing
Course takes place in the spring semester and is devoted to different video processing algorithms that include scene change detector, deinterlacer, 4D video processing, supper resolutions. Course gives intro into video processing, machine learning and deep learning. Lectures are supplemented with tasks in Python so attendees can utilize gained knowledge in practice. We are taking circa 10 students (including ones from the autumn semester). Students with the best performance have high chances of applying to the lab. More information in telegram channel @VGCourse
Computer graphics
Course program include human optical system, color and light modeling, basic image processing and analisis, basics of signal processing, rendering pipeline, OpenGL, introduction to shaders, global illumination, ray tracing and radiosity methods.
Introduction to medical image analysis
This course is devoted to application of modern deep learning algorithms to medical image analysis tasks. Introductory lecture contains information about different kinds of medical imaging technologies, as well as overview of current challenges and practical applications of medical image analysis algorithms in Russia and in the world. The main part of the course includes theoretical and practical material on convolutional neural networks, especially U-Net and its modifications, GANs, unsupervised and semi-supervised learning and also general issues of working on medical image analysis projects. The course is accompanied by the practical task where the students are free in choosing the algorithms.