Artificial Intelligence TV Generating Highlights of What You Want
A research team led by Professor Seon Joo Kim to develop AI technology that searches video recaps was selected as one of tasks for future technology by Samsung Electronics.
#1. A man is fascinated with a nature documentary series consisting of 10 episodes. In an attempt to gain favor of the woman whom he has been in love with, he wants to sort out the scenes in which flowers bloom. But it is tricky for him to watch the entire series that last over 10 hours. He simply says to the TV with an artificial intelligence technology, “Find me the scenes with blooming red flowers.”
#2. Closed circuit TV monitoring center. In its pursuit of voice phishing criminals, the police urgently requested the whereabouts of a suspect. The only clue is that a person with a blue backpack and a black cap withdrew some cash. When a police officer tells the voice recognition AI system about the features and clothes of the suspect, the system instantly points out the bank where the image of the suspect was recorded.
Associate Professor Seon Joo Kim at Yonsei University is developing AI technology which can pinpoint the image that the user wants in a long video and generate highlights. This advanced technology can understand both visual and linguistic information at the same time, and the users can make commands only with their voice without any additional devices. This technology can be applied to various fields including public safety, sports broadcasting recaps, and entertainments.
Samsung Electronics selected 40 tasks for the nurturing of future technology, including Professor Kim’s “AI Video Highlight Generation”. Since 2013, Samsung Electronics has selected the beneficiaries of the company’s support twice a year—each in the first and the second half of the year—after receiving independent research themes around the year in the fields of basic science, material engineering, and ICT.
In 2017, 18 projects were selected in the field of basic science. Samsung Electronics is spending a total of 1.5 trillion KRW in these projects for 10 years from 2013 to 2022.
Professor Seong Chan Jun
Professor Donghyun Kim