NEC technology recognizes suspicious appearance patterns in large volumes of video
Tokyo, June 12, 2018 - NEC Corporation today announced the development of a technology that automatically recognizes individuals who are displaying suspicious behaviors, such as loitering, with high precision by measuring appearance patterns (time and location, etc.) in large volumes of video data. This technology enhances "Profiling Across Spatio-Temporal Data" (Note 1), which identifies individuals who frequently appear in video footage.
NEC has developed an algorithm (method) that picks out suspicious behavior by focusing on varying appearance patterns (Figure 1). This technology statistically processes suspicious individuals' frequency of appearance in video, such as their movements and the duration of their appearances, and expresses the variations of each item in a transformation curve (Figure 2). Giving weight to the statistic items enables only individuals who show a specific appearance pattern to be picked out.
In tests of the technology using a public video data set (Note 2), it was confirmed that appearance patterns, such as loitering, waiting and walking, were correctly classified and suspicious behaviors were not overlooked.
"Ever since NEC launched the sale of the NeoFace Image data mining (Note 3) AI software utilizing Profiling Across Spatio-Temporal Data in 2016, NEC has been working on PoCs (Proof of Concept) with a number of customers. We succeeded in developing this technology by combining our know-how in analyzing the appearance patterns of people, which has been built through our demonstration experiments, with our high-speed processing technologies for large volumes of video data, which have been cultivated throughout many years," said Masayuki Mizuno, General Manager, Biometrics Research Laboratories, NEC Corporation.
Using this technology will allow users to efficiently narrow down and search for a specific target. For example, it will enable an appropriate response at an early stage for cases such as a lost or loitering individual. NEC will seek to apply this technology in security, tourism, hospitality and other businesses, aiming for it to be in practical use in FY2018.
NEC plans to present its results at the ACM International Conference on Multimedia Retrieval (ICMR2018, Period: Monday, June 11 to Thursday, June 14, 2018, Location: Yokohama, Kanagawa).
URL: http://www.icmr2018.org/
Larger viewFigure 1. Visualizing individuals' appearance patterns
Larger viewFigure 2. Transformation curves that express various appearance patterns
***
- (Note 1)Press release on November 11, 2015
NEC technology locates similar objects across multiple videos
https://www.nec.com/en/press/201511/global_20151111_02.html - (Note 2)PETS 2007 Benchmark Data.
http://www.cvg.reading.ac.uk/PETS2007/data.html - (Note 3)Press release on October 31, 2016
NEC launches AI software that searches video for specific individuals
https://www.nec.com/en/press/201610/global_20161031_04.html
About NEC Corporation
NEC Corporation is a leader in the integration of IT and network technologies that benefit businesses and people around the world. The NEC Group globally provides "Solutions for Society" that promote the safety, security, efficiency and equality of society. Under the company's corporate message of "Orchestrating a brighter world," NEC aims to help solve a wide range of challenging issues and to create new social value for the changing world of tomorrow. For more information, visit NEC at https://www.nec.com.
LinkedIn: https://www.linkedin.com/company/nec/
YouTube: https://www.youtube.com/user/NECglobalOfficial
Facebook: https://www.facebook.com/nec.global/
Twitter: https://twitter.com/NEC_corp
NEC is a registered trademark of NEC Corporation. All Rights Reserved. Other product or service marks mentioned herein are the trademarks of their respective owners. © NEC Corporation.