On The Effect Of Coding Artifacts On Acoustic Scene Classification
This addresses the problem of deploying acoustic scene classification on resource-constrained edge devices by mitigating performance degradation from audio compression, though it is incremental as it builds on prior DCASE challenge work.
The paper investigates how perceptual audio coding affects acoustic scene classification performance, finding that classification accuracy can degrade by up to 57% compared to uncompressed audio, and shows that training models with lossy compression can improve accuracy for compressed signals even with unseen codecs and bitrates.
Previous DCASE challenges contributed to an increase in the performance of acoustic scene classification systems. State-of-the-art classifiers demand significant processing capabilities and memory which is challenging for resource-constrained mobile or IoT edge devices. Thus, it is more likely to deploy these models on more powerful hardware and classify audio recordings previously uploaded (or streamed) from low-power edge devices. In such scenario, the edge device may apply perceptual audio coding to reduce the transmission data rate. This paper explores the effect of perceptual audio coding on the classification performance using a DCASE 2020 challenge contribution [1]. We found that classification accuracy can degrade by up to 57% compared to classifying original (uncompressed) audio. We further demonstrate how lossy audio compression techniques during model training can improve classification accuracy of compressed audio signals even for audio codecs and codec bitrates not included in the training process.