Abstract
Environmental sound classification (ESC) based on single-classifier-type multi-class deep neural network is gaining growing attention in current audio classification research. In order to improve the performance of multi-class deep neural networks, neural architecture search (NAS) can generally be used, but such methods often demand tremendous computational costs. This paper presents a novel multi-class framework based on local one-versus-rest (OVR) deep neural network, which can improve the classification performance of a pre-trained deep neural network at an affordable computational cost. The main idea of the framework is to first identify the weak classification category group (WeakClass group) of the pre-trained network for the actual sample using only the training data. This is achieved by using the training average confidence matrix of the pre-trained deep neural classification network. The next step is to build a sparse array of OVR subnetwork classifiers according to the WeakClass group. Afterwards, the OVR subnetwork classifiers are integrated into the original pre-trained network to form the final multi-class classifier. We apply this framework to ESC problem and the experimental results show that the proposed framework achieves a classification accuracy of 86.8% on the ESC-50 dataset, which is better than other related algorithms.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.