Divergence Based Weighting for Information Channels in Deep Convolutional Neural Networks for Bird Audio Detection

Cemre Zor; Muhammad Awais; Josef Kittler; Miroslaw Bober; Sameed Husain; Qiuqiang Kong; Christian Kroos

doi:10.1109/ICASSP.2019.8682483

Back

Conference proceeding

Divergence Based Weighting for Information Channels in Deep Convolutional Neural Networks for Bird Audio Detection

Cemre Zor, Muhammad Awais, Josef Kittler, Miroslaw Bober, Sameed Husain, Qiuqiang Kong and Christian Kroos

ICASSP 2019 - 2019 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), Vol.2019-, pp.3052-3056

ICASSP 2019 - 2019 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) (Brighton, UK, 12/05/2019 - 17/05/2019)

05/2019

DOI: https://doi.org/10.1109/ICASSP.2019.8682483

Abstract

bird audio detection

Birds

bulbul

Convergence

Convolutional neural networks

Deep convolutional neural networks

Feature extraction

KL divergence

layer initialisation

layer weighting

Task analysis

Time-frequency analysis

Training

In this paper, we address the problem of bird audio detection and propose a new convolutional neural network architecture together with a divergence based information channel weighing strategy in order to achieve improved state-of-the-art performance and faster convergence. The effectiveness of the methodology is shown on the Bird Audio Detection Challenge 2018 (Detection and Classification of Acoustic Scenes and Events Challenge, Task 3) development data set.

Metrics

38 Record Views

5 Times Cited - Web of Science

Details

Title: Divergence Based Weighting for Information Channels in Deep Convolutional Neural Networks for Bird Audio Detection
Creators: Cemre Zor - University College London
Muhammad Awais - University of Surrey, Centre for Vision, Speech & Signal Processing (CVSSP)
Josef Kittler - University of Surrey, Centre for Vision, Speech & Signal Processing (CVSSP)
Miroslaw Bober - University of Surrey, School of Computer Science and Electronic Engineering
Sameed Husain - University of Surrey, School of Computer Science and Electronic Engineering
Qiuqiang Kong - University of Surrey, Centre for Vision, Speech & Signal Processing (CVSSP)
Christian Kroos - University of Surrey, Centre for Vision, Speech & Signal Processing (CVSSP)
Publication Details: ICASSP 2019 - 2019 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), Vol.2019-, pp.3052-3056
Conference: ICASSP 2019 - 2019 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) (Brighton, UK, 12/05/2019 - 17/05/2019)
Publisher: IEEE
Date published: 05/2019
Identifiers: 99771850002346
Academic Unit: School of Computer Science and Electronic Engineering; Centre for Vision, Speech & Signal Processing (CVSSP)
Language: English
Resource Type: Conference proceeding

Divergence Based Weighting for Information Channels in Deep Convolutional Neural Networks for Bird Audio Detection

Abstract

Metrics

Details

Usage Policy