BASIC INFORMATION
Short name: SORCERER
Long name: Source Localization with Urban Acoustic Sensor Networks
Company: ASsense
Country: Belgium
Call Stage 1: F4Fp-SME-COD1 (see call details)
Proposal number: F4Fp-SME-COD200428-02
SUMMARY REMARKS & TESTBEDS
Current systems for sound source localization mainly focus on loud and impulsive events such as gunshots. Th egoal of SORCERER is to build a proof‐of‐concep t of a microphone‐based system forl ocalizing arbitrary sound source s in an urban environment. In Stage1, a pilot experiment will be conducted in the City of Things testbed in Antwerp, to gaininsightinthespatial accuracy and sensitivity that can be obtained as a function of node density. InStage2, thefocuswouldbeonimproving the sensitivity of the system using deep learning methods for detecting sounds at each node, and on implementation at the edge.
