Abstract
Audio pattern recognition is an important research topic in the machine learning area, and includes several tasks such as audio tagging, acoustic scene classification, music classification, speech emotion classification and sound event detection. Recently, neural networks have been applied to tackle the audio pattern recognition problems. However, previous systems are built on specific datasets with limited durations. Recently, in computer vision and natural language processing, systems pretrained on large-scale datasets have generalized well to several tasks. However, there is limited research on pretraining systems on large-scale audio datasets for audio pattern recognition. In this paper, we propose pretrained audio neural networks (PANNs) trained on AudioSet. Then, the PANNs are transferred to other audio related tasks. We investigate the performance and computational complexity of PANNs modeled by a variety of convolutional neural networks. Then, we propose a Wavegram architecture using both log-mel spectrogram and waveform as input feature. The best system of PANNs achieves a state-of-the-art mean average precision (mAP) of 0.439 on AudioSet tagging, outperforming the best previous system of 0.392. We transfer PANNs to six audio pattern recognition tasks, and have achieved state-of-the-art performance in several tasks. We have released the source code and pretrained models of PANNs: this https URL.
Abstract (translated)
URL
https://arxiv.org/abs/1912.10211
