MSR Vision Faculty Summit - Visual learning @ CALVIN

Опубликовано 12 августа 2016, 0:00

The dream of computer vision is to create a machine capable of human-level visual analysis. To achieve it we need powerful visual learning techniques to acquire rich models capturing the diversity of the visual world. To learn complex models and scale up to a large number of visual concepts, learning should require as little human supervision as possible. In this talk I will give an overview of recent research on visual learning at the CALVIN group, with focus on reducing the amount of human supervision necessary to learn visual concepts. A common theme is learning from images or videos labeled only by which class(es) they contain, but without information about their location in the image. Within this theme, I will present works on (a) learning object class detectors from web videos; (b) weakly supervised semantic segmentation; (c) large-scale knowledge transfer on ImageNet. vision.ee.ethz.ch/~calvin