2020/08/17 by Piotr Masztalski, Masztalski, Piotr, Mateusz Matuszewski +5 · 2 citations
Computer Science · Engineering · #Advanced Adaptive Filtering Techniques #Audio and Speech Processing (eess.AS) #FOS: Computer and information sciences #FOS: Electrical engineering #Machine Learning (cs.LG) #Music and Audio Processing #Sound (cs.SD) #Speech and Audio Processing #electronic engineering #information engineering
paper · pdf · doi:10.48550/arxiv.2008.07231
openalex publication_date 2020/08/17 · openalex created_date 2022/07/26 · openalex updated_date 2026/07/28
In this paper we introduce StoRIR - a stochastic room impulse response\ngeneration method dedicated to audio data augmentation in machine learning\napplications. This technique, in contrary to geometrical methods like\nimage-source or ray tracing, does not require prior definition of room\ngeometry, absorption coefficients or microphone and source placement and is\ndependent solely on the acoustic parameters of the room. The method is\nintuitive, easy to implement and allows to generate RIRs of very complicated\nenclosures. We show that StoRIR, when used for audio data augmentation in a\nspeech enhancement task, allows deep learning models to achieve better results\non a wide range of metrics than when using the conventional image-source\nmethod, effectively improving many of them by more than 5 %. We publish a\nPython implementation of StoRIR online\n