BEGIN:VCALENDAR VERSION:2.0 PRODID:-//wp-events-plugin.com//6.4.4//EN TZID:Asia/Jerusalem X-WR-TIMEZONE:Asia/Jerusalem BEGIN:VEVENT UID:6@cee.technion.ac.il DTSTART;TZID=Asia/Jerusalem:20210302T143000 DTEND;TZID=Asia/Jerusalem:20210302T153000 DTSTAMP:20210228T124420Z URL:https://cee.technion.ac.il/en/seminars/shai-schneider-deep-learning-mo dels-for-depth-extraction-of-landscape-scenes-from-a-single-or-a-limited-s et-of-images/ SUMMARY:Deep learning models for depth extraction of landscape scenes from a single or a limited set of images [No Categories] DESCRIPTION:Location: Civil and Environmental Engineering Speaker: Shai Sc hneider\n \n The talk will be given in Hebrew\n Abstract\n\nImage based depth extraction has been a long-studiedproblem. Classical methods rely on the geometry of two or more views observingthe same scene. Depth extracti on from a single view has also been studied\,where geometric constraints t hat required prior geometrical knowledge\, e.g.\,parallelism or known obje ct dimensions were employed. Nonetheless\, when naturalscenes are involved \, depth extraction from a single view becomes a challenge.\n\n \;\n\n To facilitate depth extraction from a single view\,this research proposes a novel neural networks based approach suited fornatural environments. To do so\, we explore the effectiveness of common lossfunctions\, and design a network suited for the problem. We also demonstrate thebi-modal nature o f the depth values in the landscape scenario\, and howutilization of this aspect improves the estimation of depth. We demonstrate ourmodel on standa rd datasets and on a new one we generated. Results showimprovement on exis ting state of the art results and predictive capabilities forup to 2500 me ters. \n Link to Zoom invitation:https://technion.zoom.us/j/98198000292 END:VEVENT BEGIN:VTIMEZONE TZID:Asia/Jerusalem X-LIC-LOCATION:Asia/Jerusalem BEGIN:STANDARD DTSTART:20201025T010000 TZOFFSETFROM:+0300 TZOFFSETTO:+0200 TZNAME:IST END:STANDARD END:VTIMEZONE END:VCALENDAR