Place: Large lecture room.
Affiliation: ETHZ vision group. Zurich, Switzerland.
Object proposals have become a key component in most of the recent state-of-the-art object detectors, as they provide a reduced set of candidate locations for objects in the image. They allow, therefore, powerful machine learning algorithms (aka CNNs) to be applied in these scenarios. In this talk I will first present Multiscale Combinatorial Grouping (MCG) , a state-of-the-art object proposal and hierarchical segmentation technique. Then, I will describe some possible future lines of research in object proposal techniques by highlighting the new challenges faced by going from Pascal to COCO .
bio: Jordi Pont-Tuset is a post-doctoral researcher in Prof. Luc Van Gool’s ETHZ vision group. Previously, he did an 8-month internship at Disney Research Zürich, under the supervision of Prof. Aljoscha Smolic, he visited Prof. Jitendra Malik’s vision group in UC Berkeley, and he collaborated with the startup Fezoo. He is a mathematician, engineer, and PhD in computer vision by UPC Barcelonatech under the supervision of Prof. Ferran Marques.