{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/sparse-gaussian-process-audio-source","title":"Sparse Gaussian Process Audio Source Separation Using Spectrum Priors in the Time-Domain","arxiv_id":"1810.12679","date":"2018-10-30","proceeding":null,"authors":["Pablo A. Alvarado","Mauricio A. Álvarez","Dan Stowell"],"abstract":"Gaussian process (GP) audio source separation is a time-domain approach that\ncircumvents the inherent phase approximation issue of spectrogram based\nmethods. Furthermore, through its kernel, GPs elegantly incorporate prior\nknowledge about the sources into the separation model. Despite these compelling\nadvantages, the computational complexity of GP inference scales cubically with\nthe number of audio samples. As a result, source separation GP models have been\nrestricted to the analysis of short audio frames. We introduce an efficient\napplication of GPs to time-domain audio source separation, without compromising\nperformance. For this purpose, we used GP regression, together with spectral\nmixture kernels, and variational sparse GPs. We compared our method with\nLD-PSDTF (positive semi-definite tensor factorization), KL-NMF\n(Kullback-Leibler non-negative matrix factorization), and IS-NMF (Itakura-Saito\nNMF). Results show that the proposed method outperforms these techniques.","url_abs":"http://arxiv.org/abs/1810.12679v3","url_pdf":"http://arxiv.org/pdf/1810.12679v3.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"sparse-gaussian-process-audio-source","repo_url":"https://github.com/PabloAlvarado/ssgp","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"tf","reach":null}],"tasks":[{"task_slug":"audio-source-separation","task_name":"Audio Source Separation"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}