UTDSIDSIEventsDSI Meet Up: Seminar Series on Computer Vision

DSI Meet Up: Seminar Series on Computer Vision Pure motion for video analysis

This DSI Meet Up focuses on Pure Motion for Video Analysis, exploring how motion based representations can improve the way computer vision models analyse and understand video. The series is co-organised by members of the DMB group and collaborating colleagues across the University of Twente.

The “Computer Vision” seminar series aims to bring together students and researchers passionate about enabling machines to interpret and understand visual information. Our goal is to highlight cutting-edge research, showcase innovative applications, and foster collaborations among experts from nearby institutions. Each session will feature an invited speaker presenting their work on a specific topic within the broad and fast-evolving field of computer vision.

By joining these seminars, you’ll gain insights into the latest academic and industrial developments, exchange ideas with peers, and explore opportunities for joint research and innovation. Whether you are a PhD candidate, an experienced researcher, or simply curious about computer vision, feel welcome to join these seminars.

The seminars are open to everyone at UT interested in the topic. 

Abstract

Pure motion for video analysis

Videos involve motion. Currently, computer vision models for video analysis tie the motion information to the appearance of the objects and scenes. This leads to models that cannot generalize well to new settings. I argue that video analysis should focus on motion: e.g. to recognise the action of "chopping vegetables", we can ignore the kitchen appearance, the shape of the knife or the colour of the cutting board, and focus on the motion of the hands. My work researches the utility of "pure motion representations" as encoded by Lagrangian and Eulerian representations, evaluated on benchmarks where appearance is uninformative. My aim is to enable a research shift towards strong motion representations, reusable models, and data savings.

Speaker

Silvia Pintea

Silvia Pintea is a Computer Vision researcher in the Deep learning for Perception Lab at Tilburg University. 

Previously, she worked at LUMC (Division of image processing) and at Delft University of Technology (Computer vision lab, PRB). Before, she was an R&D Engineer at Blippar on augmented reality. Silvia obtained her PhD in Computer Vision at the University of Amsterdam in 2017.

Next to puzzling about visual learning problems: from colours, edges and corners in images, to time and motion in videos, Silvia is also a passionate climber and snowboarder. 

Registration & contact 

Please register via the registration button below. For more information and questions regarding the seminar series on computer vision, please contact Estefanía Talavera Martínez.

DSI Meet Up: Seminar Series on Computer Vision Pure motion for video analysis
Register