Finding a person with a Wi-Fi device in a crowd of pedestrians

interests include human-robot interaction, interactive humanoid robots, and networked robots. Kaoru Kurumizawa received a master's degree in engineering from Osaka University in 2011 and is currently working at Toyota. His research interests include networked robots, human-robot interfaces, and field trials. Takayuki Kanda received a Ph.D. in computer science from Kyoto University in 2003. He is currently a senior researcher at ATR-IRC. His research interests include intelligent robotics, human-robot interaction, and vision-based mobile robots. leader in the ATR Hiroshi Ishiguro Laboratories. Norihiro Hagita received a Ph.D. in electrical engineering from Keio University in 1986. He is now a director of ATR-IRC. His research interests include communication robots, networked robot systems, interaction media, and pattern recognition. This paper proposes a technique to find and track persons carrying Wi-Fi devices. Our method integrates Wi-Fi signal strength information with LRF-based people-tracking results and estimates device holders from the statistical similarity of a time series of observed Wi-Fi signal strengths and predicted signal strengths based on the positions provided by people-tracking. Since human bodies significantly hinder radio transmission, our algorithm considers the Wi-Fi device location of the holding person. Experiment result revealed that our developed algorithm successfully found the device's holder with 87.2% successful identification rate in a real environment.