Vabljeni na naslednjo debato Katedre za umetno inteligenco, ki bo v četrtek, 10. marca 2022, ob 14.15 v predavalnici 19. Debato z naslovom "A discriminative segmentation method for visual object tracking" je pripravil Alan Lukežič, raziskovalec v Laboratoriju za umetne vizualne spoznavne sisteme UL FRI.

Naslov: A discriminative segmentation method for visual object tracking

Opis: Visual object tracking remains one of the fundamental problems in computer vision, which progressed significantly in the last decade. The progress has been facilitated by the advent of deep learning as well as by emergence of moderated datasets and challenges. Traditional trackers reported target location in form of bounding boxes, which is in many cases not very accurate approximation. To address this issue, tracking has started to converge with a closely related field of video object segmentation in the last two years. We are now witnessing emergence of trackers that segment the target, while by far outperforming their bounding-box-oriented predecessors. In this talk, I will present our segmentation-based tracker, which was recently presented at the CVPR conference and the extended version at IEEE TPAMI.

O predavatelju: Alan Lukežič is a researcher in a Visual cognitive systems laboratory. He finished a PhD in visual object tracking with discriminative correlation methods in May 2021. Currently he is still working on closely related research topics.