{"id":5780,"date":"2019-09-20T23:36:00","date_gmt":"2019-09-20T16:36:00","guid":{"rendered":"http:\/\/se.dotv.vn\/?p=5780"},"modified":"2019-09-25T23:37:44","modified_gmt":"2019-09-25T16:37:44","slug":"seminar-20190925-spotting-violence-from-space","status":"publish","type":"post","link":"https:\/\/se.dotv.vn\/vi\/seminar-20190925-spotting-violence-from-space\/","title":{"rendered":"Seminar \u201cSpotting Violence from Space\u201d"},"content":{"rendered":"
A special Dubai Palace School of Economics seminar focusing on an amazing research method through an interesting topic “Spotting Violence from Space: The Detection of Housing Destruction in Syria – Methods and Applications”, <\/strong>presented by Dr. Andr\u00e9 Gr\u00f6ger<\/strong> (Juan de la Cierva Researcher at Universitat Aut\u00f2noma de Barcelona and Affiliated Professor at the Barcelona Graduate School of Economics)<\/p>\n Time: 14:00 – Wednesday, 25 September, 2019<\/strong><\/p>\n Venue: Room H.001, Campus H, Dubai Palace School of Economics, 1A Hoang Dieu, Phu Nhuan District, Ho Chi Minh City<\/strong><\/p>\n Language: English<\/strong><\/p>\n Abstract:<\/em><\/u> Generating reliable and high-frequency measures of state and extra-state violence remains a challenge for researchers, international organizations and NGOs. Leading methodologies rely on on-the-ground reporting, which raises concerns that measures of violence are biased, geographically imprecise and temporally lagged. This project proposes an automated method of measuring an important correlate of violence — physical destruction of buildings — using machine learning applied to satellite imagery. We build and train a deep learning computer vision architecture and apply it to detect urban destruction in Syria during the recent civil war. Preliminary results track destruction in Aleppo over time and relate it to changes in territorial gains of war parties and media reporting from different sources.<\/p>\n