Marine mammal recognition is very important in marine environmental protection and marine resources development. In order to solve the problem of over fitting in marine mammal sound recognition task, this paper proposes a diversity based on Convolutional Neural Network (CNN) ensemble regularization method for marine mammal sound recognition, which integrates three kinds of deep CNN models. Firstly, the Bagging algorithm is used to generate multiple recognition models, then a subset of recognition models is selected based on four diversity measures, and finally the recognition models of each subset are integrated by the majority voting method. The experimental results show that the ensemble model obtained by the selective ensemble method improves the accuracy compared with the single recognition model, and improves the stability of the recognition results and the generalization ability compared with the random selection ensemble method.
CNN Ensemble Regularization Method Based on Diversity Measure for Marine Mammals Sound Recognition
Lect. Notes Electrical Eng.
International Conference on Autonomous Unmanned Systems ; 2021 ; Changsha, China September 24, 2021 - September 26, 2021
Proceedings of 2021 International Conference on Autonomous Unmanned Systems (ICAUS 2021) ; Kapitel : 259 ; 2637-2647
2022-03-18
11 pages
Aufsatz/Kapitel (Buch)
Elektronische Ressource
Englisch
British Library Conference Proceedings | 1999
|British Library Conference Proceedings | 2001
British Library Conference Proceedings | 1995
|UAV-based marine mammals positioning and tracking system
Tema Archiv | 2011
|Radar Detection of Marine Mammals
NTIS | 2009
|