A team of University of Illinois at Urbana-Champaign and Inspirit IoT engineers has won the Design Automation Conference (DAC) 2017 International Hardware Design competition with their Low Power, Low Cost Audio Based Security IoT System. The team leveraged a low-power, inexpensive Lattice FPGA together with a custom-built audio capture board together with custom machine learning in order to detect screaming and gunshots and perform direction finding of audio events on a low-power IoT system. The system was demonstrated as a stand-alone system for detecting audio events and could also be paired with a video surveillance system to guide video surveillance through pan and zoom, more intelligent power consumption than always-on video systems, and more intelligent video capture and communications with video-based machine learning.
A description of their winning system design appears in their short paper
Low Power, Low Cost Audio Based Security IoT System
iSmart Team: University of Illinois at Urbana-Champaign (Di He, Dushyant Bhan, Zizhen Liu, Xinheng Liu, Deming Chen). Inspirit IoT, Inc. (Zuofu Cheng, Swathi Gurumani, Kyle Rupnow)
We just sent you an email. Please click the link in the email to confirm your subscription!