Startseite // SnT // News & E... // Best Paper Award for mixed team from SnT and IEE

Best Paper Award for mixed team from SnT and IEE

twitter linkedin facebook email this page
Veröffentlicht am Donnerstag, den 18. Juni 2015

On June 12th , SnT-researchers Kassem Al Ismaeil and Dr. Djamila Aouada together with SnT-director Prof. Björn Ottersten as well as Thomas Solignac and Dr. Bruno Mirbach – both from IEE, a global leader in automotive safety sensing systems – have received the Best Paper Award of MSF 2015.

MSF is the 2nd IEEE/ISPRS Joint Workshop on Multi-Sensor Fusion for Dynamic Scene Understanding as part of the premier annual Computer Vision event. It comprises the main IEEE Conference in Computer Vision and Pattern Recognition (CVPR) and several co-located workshops and courses. In their paper the authors propose to enhance low resolution dynamic depth videos containing freely non–rigidly moving objects with a new dynamic multi–frame super–resolution algorithm.

Existing methods are either limited to rigid objects, or restricted to global lateral motions discarding radial displacements. The team addresses these shortcomings by accounting for non–rigid displacements in 3D. In addition to 2D optical flow, the scientists estimate the depth displacement, and simultaneously correct the depth measurement by so called Kalman filtering. This concept is incorporated efficiently in a multi-frame super-resolution framework. It is formulated in a recursive manner that ensures an efficient deployment in real-time.

Results show the overall improved performance of the proposed method as compared to alternative approaches, and specifically in handling relatively large 3D motions. Test examples range from a full moving human body to a highly dynamic facial video with varying expressions.

As the fusion of dynamic 2D/3D multi-sensor data has not been extensively explored in the vision community, the MSF 2015 workshop´s aim was to encourage interdisciplinary interaction and collaboration from computer vision, remote sensing, robotics and photogrammetry communities. Held in Boston, Massachuchsetts (USA), it served as a forum for research groups from academia and industry.