1 October 2019 in APL Materials
YKYoshiharu Krockenberger
Materials informatics exploiting machine learning techniques, e.g., Bayesian optimization (BO), have the potential to reduce the number of thin-film growth runs for optimization of thin-film growth conditions through incremental updates of machine learning models in accordance with newly measured data. Here, we demonstrated BO-based molecular beam epitaxy (MBE) of SrRuO3, one of the most intensively studied materials in the research field of oxide electronics, mainly owing to its unique nature as a ferromagnetic metal. To simplify the intricate search space of entangled growth conditions, we ran the BO for a single condition while keeping the other conditions fixed. As a result, high-crystalline-quality SrRuO3 film exhibiting a high residual resistivity ratio of over 50 as well as strong perpendicular magnetic anisotropy was developed in only 24 MBE growth runs in which the Ru flux rate, growth temperature, and O3-nozzle-to-substrate distance were optimized. Our BO-based search method provides an efficient experimental design that is not as dependent on the experience and skills of individual researchers, and it reduces experimental time and cost, which will accelerate materials research.
1 January 2019 in APSIPA Transactions on Signal and Information Processing
14 May 2021 in IEEE/ACM Transactions on Audio, Speech, and Language Processing
Blind source separation (BSS) of audio signals aims to separate source signals from their mixtures recorded by microphones. The applications include automatic speech recognition in a noisy/multi-speaker environment, hearing aids, and music analysis. Independent component analysis (ICA) can perform BSS efficiently, but it is basically inapplicable to the underdetermined casethe number of sources the number of microphones. In contrast, a BSS approach using the multichannel Wiener filter (MWF) is applicable even to the underdetermined case, but conventional BSS methods based on this approachincluding full-rank spatial covariance analysis (FCA)are highly inefficient. This is because, in these methods, massive numbers of matrix inversions are required for designing the MWF. To obtain the best of both worlds, we take a joint diagonalization approach: We restrict spatial covariance matrices of all sources to the class of jointly diagonalizable matrices. This enables the above matrix inversions to be replaced by mere scalar inversions of the diagonal elements. Based on this, we present FastFCA and FastMNMF, efficient methods for underdetermined BSS. In experiments, FastFCA was several orders of magnitude faster than FCA without sacrificing separation performance. We also present a unified framework for underdetermined and determined BSS, which highlights theoretical connections between various methods including FastFCA and FastMNMF. The efficiency of our methods makes them suitable for large data (e.g., data augmentation for machine learning) or limited computational resources encountered in, e.g., hearing aids, distributed microphone arrays, and online BSS.
11 December 2013 in IEEE/ACM Transactions on Audio, Speech, and Language Processing
Sound source localization and separation from a mixture of sounds are essential functions for computational auditory scene analysis. The main challenges are designing a unified framework for joint optimization and estimating the sound sources under auditory uncertainties such as reverberation or unknown number of sounds. Since sound source localization and separation are mutually dependent, their simultaneous estimation is required for better and more robust performance. A unified model is presented for sound source localization and separation based on Bayesian nonparametrics. Experiments using simulated and recorded audio mixtures show that a method based on this model achieves state-of-the-art sound source separation quality and has more robust performance on the source number estimation under reverberant environments.
1 March 2022 in IEEE/ACM Transactions on Audio, Speech, and Language Processing
This paper develops a framework that can accurately perform denoising, dereverberation, and source separation using a relatively small number of microphones. It has been empirically confirmed that Independent Vector Analysis (IVA) can blindly separate N sources from their sound mixture even with diffuse noise when a sufficiently large number (= M) of microphones are available (i.e., M N). However, the estimation accuracy is seriously degraded when the number of microphones, or more specifically M N (0), decreases. To overcome this IVA limitation, we propose switching IVA (swIVA) in this paper. With swIVA, the time frames of an observed signal with time-varying characteristics are clustered into several groups, each of which can be well handled by IVA with a small number of microphones, and thus accurate estimation can be achieved by individually applying IVA to each group. Conventionally, a switching mechanism was introduced into a Minimum-Variance Distortionless Response (MVDR) beamformer, and this paper extends the mechanism to work with a blind source separation algorithm. To incorporate dereverberation capability, we further extend swIVA to a blind Convolutional beamforming algorithm (swCIVA) that integrates swIVA and switching Weighted Prediction Error-based dereverberation (swWPE) in a jointly optimal way. With swCIVA, two different time-varying characteristics of an observed signal are captured for dereverberation and source separation to achieve effective estimation. We show that both swIVA and swCIVA can be optimized effectively based on blind signal processing, and their performance can be further improved using a spatial guide for initialization. Experiments demonstrate that both the proposed methods largely outperformed conventional IVA and its convolutional beamforming extension (CIVA) in terms of objective signal quality and automatic speech recognition scores when using relatively few microphones.