Audio Scene Classification with Deep Recurrent Neural Networks
This work addresses audio scene classification for applications like surveillance or smart devices, representing an incremental improvement over existing methods.
The paper tackles audio scene classification by transforming audio into label tree embeddings and using a deep GRU-based RNN for sequence-to-label classification, achieving an F1-score of 97.7% on the LITIS Rouen dataset and reducing relative classification error by 35.3% compared to prior results.
We introduce in this work an efficient approach for audio scene classification using deep recurrent neural networks. An audio scene is firstly transformed into a sequence of high-level label tree embedding feature vectors. The vector sequence is then divided into multiple subsequences on which a deep GRU-based recurrent neural network is trained for sequence-to-label classification. The global predicted label for the entire sequence is finally obtained via aggregation of subsequence classification outputs. We will show that our approach obtains an F1-score of 97.7% on the LITIS Rouen dataset, which is the largest dataset publicly available for the task. Compared to the best previously reported result on the dataset, our approach is able to reduce the relative classification error by 35.3%.