Environmental Sound Classification Using CNN Based on Mel-spectogram
Zhihao Gao, Tongyang Liu, Ming Zhu, Jing Li, Yingying Ning, Zhiqiang Wang
Abstract
In recent years, with the development of deep learning technology, the combination of deep neural network and audio data processing and analysis has become a new research hotspot, especially the representative convolutional neural network has achieved significant results on sound classification tasks. Therefore, in this paper, we propose a CNN structure, combined with extracting Mel-spectogram features and using data enhancement to increase the training data, and finally compare the classification accuracy under different CNN networks, and the experiments show that the method in this paper has a higher accuracy.
Topics & Concepts
Computer scienceConvolutional neural networkArtificial intelligenceDeep learningArtificial neural networkPattern recognition (psychology)Hotspot (geology)Speech recognitionMachine learningGeophysicsGeologyMusic and Audio ProcessingSpeech and Audio ProcessingMusic Technology and Sound Studies