Learning audio sequence representations for acoustic event classification
作者:
Highlights:
• The proposed method learns higher representations independent of audio length.
• The proposed method was validated by a large-scale acoustic event database.
• It can be easily transferred to fulfil other audio or multi-media related studies.
摘要
•The proposed method learns higher representations independent of audio length.•The proposed method was validated by a large-scale acoustic event database.•It can be easily transferred to fulfil other audio or multi-media related studies.
论文关键词:Audio sequence-to-vector,Recurrent autoencoder,Acoustic event classification,Machine learning,Deep learning,Computer audition
论文评审过程:Received 6 May 2020, Revised 1 March 2021, Accepted 3 April 2021, Available online 20 April 2021, Version of Record 24 April 2021.
论文官网地址:https://doi.org/10.1016/j.eswa.2021.115007