%0 Journal Article %1 attention-based-crnn %A Zhang, Zhichao %A Xu, Shugong %A Zhang, Shunqing %A Qiao, Tianhao %A Cao, Shan %D 2021 %J Neurocomputing %K attention_mechanism audio_classification convolutional_recurrent_neural_network thema:cnn_and_attention_methods_for_audio_classification %P 896-903 %R https://doi.org/10.1016/j.neucom.2020.08.069 %T Attention based convolutional recurrent neural network for environmental sound classification %U https://www.sciencedirect.com/science/article/pii/S0925231220313618 %V 453 %X Environmental sound classification (ESC) is a challenging problem due to the complexity of sounds. The classification performance is heavily dependent on the effectiveness of representative features extracted from the environmental sounds. However, ESC often suffers from the semantically irrelevant frames and silent frames. In order to deal with this, we employ a frame-level attention model to focus on the semantically relevant frames and salient frames. Specifically, we first propose a convolutional recurrent neural network to learn spectro-temporal features and temporal correlations. Then, we extend our convolutional RNN model with a frame-level attention mechanism to learn discriminative feature representations for ESC. We investigated the classification performance when using different attention scaling function and applying different layers. Experiments were conducted on ESC-50 and ESC-10 datasets. Experimental results demonstrated the effectiveness of the proposed method and our method achieved the state-of-the-art or competitive classification accuracy with lower computational complexity. We also visualized our attention results and observed that the proposed attention mechanism was able to lead the network tofocus on the semantically relevant parts of environmental sounds.

@article{attention-based-crnn, abstract = {Environmental sound classification (ESC) is a challenging problem due to the complexity of sounds. The classification performance is heavily dependent on the effectiveness of representative features extracted from the environmental sounds. However, ESC often suffers from the semantically irrelevant frames and silent frames. In order to deal with this, we employ a frame-level attention model to focus on the semantically relevant frames and salient frames. Specifically, we first propose a convolutional recurrent neural network to learn spectro-temporal features and temporal correlations. Then, we extend our convolutional RNN model with a frame-level attention mechanism to learn discriminative feature representations for ESC. We investigated the classification performance when using different attention scaling function and applying different layers. Experiments were conducted on ESC-50 and ESC-10 datasets. Experimental results demonstrated the effectiveness of the proposed method and our method achieved the state-of-the-art or competitive classification accuracy with lower computational complexity. We also visualized our attention results and observed that the proposed attention mechanism was able to lead the network tofocus on the semantically relevant parts of environmental sounds.}, added-at = {2022-07-12T01:34:06.000+0200}, author = {Zhang, Zhichao and Xu, Shugong and Zhang, Shunqing and Qiao, Tianhao and Cao, Shan}, biburl = {https://www.bibsonomy.org/bibtex/2da93181010963739d62d39ce11154b05/fachter}, doi = {https://doi.org/10.1016/j.neucom.2020.08.069}, interhash = {4b700231319dee6ac55e38bc13a1178d}, intrahash = {da93181010963739d62d39ce11154b05}, issn = {0925-2312}, journal = {Neurocomputing}, keywords = {attention_mechanism audio_classification convolutional_recurrent_neural_network thema:cnn_and_attention_methods_for_audio_classification}, pages = {896-903}, timestamp = {2022-07-12T10:20:36.000+0200}, title = {Attention based convolutional recurrent neural network for environmental sound classification}, url = {https://www.sciencedirect.com/science/article/pii/S0925231220313618}, volume = 453, year = 2021 }