IARPA Issues Two Technology Challenges
Researchers seek tools to measure credibility and to monitor videos.
The Intelligence Advanced Research Projects Activity (IARPA) has announced two new challenges: the Credibility Assessment Standardized Evaluation (CASE) Challenge, which seeks methods for measuring the performance of credibility assessment techniques and technologies, and the Activities in Extended Video (ActEV) Prize Challenge, which aims to develop algorithms that will monitor surveillance videos for suspicious activity.
Regarding the CASE Challenge, IARPA officials point out in the announcement that there are no standardized and rigorous tests to evaluate how accurate credibility assessment tools really are. “Countless studies have tested a variety of credibility assessment techniques and have attempted to use them to rigorously determine when a source and/or a message is credible and, more specifically, when a person is lying or telling the truth,” the announcement states. “Despite the large and lengthy investment in such research, a rigorous set of valid methods that are useful in determining the credibility of a source or their information across different applications remains difficult to achieve.”
The announcement for the ActEV Prize Challenge invites participants from across the globe and across disciplines—with or without experience in video analysis—to create algorithms that detect and localize activities in video.
The ActEV Prize Challenge will be conducted in two stages: an open leaderboard evaluation during which all eligible solutions will be evaluated, and an independent evaluation stage in which the top eight solutions will be tested on sequestered data by the National Institute for Standards and Technology.
The announcement explains why video monitoring tools are needed: “The volume of video data collected from security cameras has grown dramatically in recent years. However, there has not been a commensurate increase in the capabilities of automated analytics for real-time alerting, triaging or forensic analysis of video. Operators of camera networks are typically overwhelmed with the volume of video they must monitor, and cannot afford to view or analyze even a small fraction of their video footage,” the announcement states. “Automated methods that identify and localize activities in extended video are necessary to alleviate the current, manual process of monitoring by human operators.”