2020/10/21 by An Tran, Tran, An, Konstantinos Drossos +3
Computer Science · Engineering · #Audio and Speech Processing (eess.AS) #FOS: Computer and information sciences #FOS: Electrical engineering #Machine Learning (cs.LG) #Multimedia (cs.MM) #Sound (cs.SD) #cs.LG #cs.MM #cs.SD #eess.AS #electronic engineering #information engineering
paper · pdf · doi:10.48550/arxiv.2010.11098
Submitted for review at ICASSP2021
arxiv created 2020/10/21 · arxiv updated 2020/10/22
Automated audio captioning (AAC) is a novel task, where a method takes as an input an audio sample and outputs a textual description (i.e. a caption) of its contents. Most AAC methods are adapted from from image captioning of machine translation fields. In this work we present a novel AAC novel method, explicitly focused on the exploitation of the temporal and time-frequency patterns in audio. We employ three learnable processes for audio encoding, two for extracting the local and temporal information, and one to merge the output of the previous two processes. To generate the caption, we employ the widely used Transformer decoder. We assess our method utilizing the freely available splits of Clotho dataset. Our results increase previously reported highest SPIDEr to 17.3, from 16.2.