Natural Environment Protection

Forest EARS: Forest Early Detection of Human Activity using Audio SuRveillance Systems

Description

Anomalous sound detection offers new possibilities for enhancing surveillance methods in general, and particularly monitoring illegal deforestation activities, such as tree cutting, illegal logging and poaching. Although video surveillance has achieved a high level of accuracy , audio data can still be more effective in some situations than video, most because of its omnidirection and insensitivity to luminosity conditions. In order to identify illegal deforestation activities, such as tree logging, this research proposed two ASD methods based on Graph Neural Networks that aim to detect sawchain sounds in a natural landscape.

Collaborators

Bradford University La Rochelle University

Zied MNASRI (Associate Professor)                                 Thierry BOUWMANS