In this paper, we propose the application of denoising to microphone classification, to enable its usage on content with unfavorable noisy conditions. We first describe the proposed integrated approach; afterwards we discuss the baseline algorithm for microphone classification, and the various denoising procedures which can be combined with it in the time or spectral domain; lastly, we determine the best performing denoising procedure, and evaluate the performance of the integrated approach with several SNR levels of additive input noise. In comparison to the reference baseline, the proposed method achieves an average accuracy increase of about 25% on denoised content.
View on arXiv