著者

Shinichi MOGAMI Yoshiki MITSUI Norihiro TAKAMUNE Daichi KITAMURA Hiroshi SARUWATARI Yu TAKAHASHI Kazunobu KONDO Hiroaki NAKAJIMA Hirokazu KAMEOKA

出版者

The Institute of Electronics, Information and Communication Engineers

雑誌

IEICE TRANSACTIONS on Fundamentals of Electronics, Communications and Computer Sciences (ISSN:09168508)

巻号頁・発行日

vol.E102-A, no.2, pp.458-463, 2019-02-01

被引用文献数

5

---

In this letter, we propose a new blind source separation method, independent low-rank matrix analysis based on generalized Kullback-Leibler divergence. This method assumes a time-frequency-varying complex Poisson distribution as the source generative model, which yields convex optimization in the spectrogram estimation. The experimental evaluation confirms the proposed method's efficacy.

著者

Naoya Murashima Hirokazu Kameoka Li Li Shogo Seki Shoji Makino

出版者

Research Institute of Signal Processing, Japan

雑誌

Journal of Signal Processing (ISSN:13426230)

巻号頁・発行日

vol.25, no.4, pp.145-149, 2021-07-01 (Released:2021-07-01)

参考文献数

24

---

This paper deals with single-channel speaker-dependent speech separation. While discriminative approaches using deep neural networks (DNNs) have recently proved powerful, generative approaches, including methods based on non-negative matrix factorization (NMF), are still attractive because of their flexibility in handling the mismatch between training and test conditions. Although NMF-based methods work reasonably well for particular sound sources, one limitation is that they can fail to work for sources with spectrograms that do not comply with the NMF model. To address this problem, attempts have recently been made to replace the NMF model with DNNs. With a similar motivation to these attempts, we propose in this paper a variational autoencoder (VAE)-based monaural source separation (VASS) method using a conditional VAE (CVAE) for source spectrogram modeling. We further propose an extension of the VASS method, called the discriminative VASS (DVASS) method, which uses a discriminative criterion for model training so that the separated signals directly become optimal. Experimental results revealed that the VASS method performed better than an NMF-based method, and the DVASS method performed better than the VASS method.

著者

Akisato Kimura Masashi Sugiyama Hitoshi Sakano Hirokazu Kameoka

雑誌

情報処理学会論文誌数理モデル化と応用(TOM) (ISSN:18827780)

巻号頁・発行日

vol.6, no.1, pp.136-145, 2013-03-12

---

It is well known that dimensionality reduction based on multivariate analysis methods and their kernelized extensions can be formulated as generalized eigenvalue problems of scatter matrices, Gram matrices or their augmented matrices. This paper provides a generic and theoretical framework of multivariate analysis introducing a new expression for scatter matrices and Gram matrices, called Generalized Pairwise Expression (GPE). This expression is quite compact but highly powerful. The framework includes not only (1) the traditional multivariate analysis methods but also (2) several regularization techniques, (3) localization techniques, (4) clustering methods based on generalized eigenvalue problems, and (5) their semi-supervised extensions. This paper also presents a methodology for designing a desired multivariate analysis method from the proposed framework. The methodology is quite simple: adopting the above mentioned special cases as templates, and generating a new method by combining these templates appropriately. Through this methodology, we can freely design various tailor-made methods for specific purposes or domains.

著者

Akisato Kimura Masashi Sugiyama Takuho Nakano Hirokazu Kameoka Hitoshi Sakano Eisaku Maeda Katsuhiko Ishiguro

雑誌

情報処理学会論文誌数理モデル化と応用(TOM) (ISSN:18827780)

巻号頁・発行日

vol.6, no.1, pp.128-135, 2013-03-12

---

Canonical correlation analysis (CCA) is a powerful tool for analyzing multi-dimensional paired data. However, CCA tends to perform poorly when the number of paired samples is limited, which is often the case in practice. To cope with this problem, we propose a semi-supervised variant of CCA named SemiCCA that allows us to incorporate additional unpaired samples for mitigating overfitting. Advantages of the proposed method over previously proposed methods are its computational efficiency and intuitive operationality: it smoothly bridges the generalized eigenvalue problems of CCA and principal component analysis (PCA), and thus its solution can be computed efficiently just by solving a single eigenvalue problem as the original CCA.