Title :
The acoustic adaptive frightening device-framework and algorithms
Author :
Steen, Kim Arild ; Karstoft, Henrik
Author_Institution :
Dept. of Eng., Aarhus Univ., Aarhus, Denmark
Abstract :
Current frightening devices are inefficient, because of habituation. This paper presents concepts and algorithms for an adaptive frightening device, which is capable of measuring its own success and thereby alter frightening strategies if beginning habituation is observed. The framework includes acoustic detection and recognition of conflict species and their behavior. The proposed framework has been tested in field experiments where goose activity was heavily reduced for 5-6 weeks.
Keywords :
acoustic devices; acoustic signal detection; bioacoustics; acoustic adaptive frightening device; acoustic detection; acoustic recognition; conflict species; frightening strategy; Acoustics; Birds; Pattern recognition; Signal processing algorithms; Signal to noise ratio; GFCC; GMM; MFCC; SVM; intelligent agent; pattern recognition; wildlife management;
Conference_Titel :
Signal and Information Processing (GlobalSIP), 2014 IEEE Global Conference on
Conference_Location :
Atlanta, GA
DOI :
10.1109/GlobalSIP.2014.7032295