Sungmok Hwang

Interpretations on principal components analysis of head-related impulse responses in the median planea)

A principal components analysis of the median-plane head-related impulse responses (HRIRs) in the CIPIC HRTF database reveals that the individual HRIRs can be reconstructed by a linear combination of 12 principal components (PCs) within 5% of error in the least-squares sense. The PCs include the intersubject and interelevation variations in the median-plane HRIRs. Each PC provides sound cues for the front-back discrimination and/or the vertical perception. There exist common systematic elevation dependencies in the weights of lower-numbered PCs which contribute to the pinna/head diffractions, whereas the elevation dependencies in the weights of higher-numbered PCs are different from subject to subject.

Sound direction estimation using artificial ear

We propose a novel design of artificial robot ear and sound direction estimation method using the measured two output signals only. The spectral features in head-related transfer functions and in interaural transfer functions are distinctive across voice frequency band. Thus, these features provide effective sound cues to estimate sound direction using the measured two output signals. Bilateral asymmetry of microphone positions can enhance the estimation performance even in the median plane where interaural differences are vanished. Sound direction is estimated from interaural time difference and correlation between the log magnitudes of interaural transfer functions. The feasibility and the estimation performance of the designed artificial ear and the estimation method are verified in a real environment. In the experiment, we confirm that robots with the proposed artificial ear can find the direction of user from two output signals only with reasonable accuracy.

Artificial Robot Ear Design for Sound Direction Estimation

We propose a novel design of artificial robot ear for sound direction estimation using measured two outputs only. The spectral features in head-related transfer functions and in interaural transfer functions (ITFs) are distinctive in the voice frequency band. Thus, these features provide effective sound cues to estimate the sound direction using two measured ear outputs only without input information. Especially, the direction of median source can be estimated by the bilateral asymmetry of microphone positions, whereas many previous studies, which depend on the interaural differences, cannot achieve. We propose a localization method to estimate the lateral and vertical angles simultaneously. The lateral angle is estimated using interaural time delay and Woodworth and Schlosbergs formula, and the front-back discrimination is achieved by the presence of the spectral features in ITF estimated from two measured outputs. The vertical angle of the frontal source is estimated by comparing the spectral features in the estimated ITF with those in the database built in an anechoic chamber.

Sound source localization in median plane using artificial ear

Sound source localization is the method using the measurements of the acoustic signals from microphone arrays in acoustical engineering. This technique has been used broadly in 3-D sound technology, humanoid robot and teleconferencing and so on. For robot industry, their ultimate purpose is to be with human being. This is why the industry is demanding applicable robotpsilas auditory system in the form of artificial ears like humanpsilas external ear such as ear pinna. It has more benefits to make use of auditory system with ear pinna to humanoid robots for HRI. In this paper, we propose a specific sound source localization method using a pair of artificial ears, each of which consisting of a single ear pinna and two microphones. The feasibility and localization performance of proposed method for speech signal in median plane is shown. Through the experiment in office environment, we confirm that robots with artificial ears can estimate the elevation angle of speech signal just using two microphone output signals.

Comparison of Head-related Transfer Function Models Based on Principal Components Analysis

This study deals with modeling of head-related transfer functions(HRTFs) using principal components analysis(PCA) in the time and frequency domains. Four PCA models based on head-related impulse responses(HRIRs), complex-valued HRTFs, augmented HRTFs, and log-magnitudes of HRTFs are investigated. The objective of this study is to compare modeling performances of the PCA models in the least-squares sense and to show the theoretical relationship between the PCA models. In terms of the number of principal components needed for modeling, the PCA model based on HRIR or augmented HRTFs showed more efficient modeling performance than the PCA model based on complex-valued HRTFs. The PCA model based on HRIRs in the time domain and that based on augmented HRTFs in the frequency domain are shown to be theoretically equivalent. Modeling performance of the PCA model based on log-magnitudes of HRTFs cannot be compared with that of other PCA models because the PCA model deals with log-scaled magnitude components only, whereas the other PCA models consider both magnitude and phase components in linear scale.

Modeling and customization of head-related transfer functions using principal component analysis

This study deals with modeling and customization of head-related transfer functions (HRTFs) in the median plane based on a weighted linear summation of a set of general basis functions obtained from a specific huge HRTF database. The 12 principal components (PCs) were extracted from principal components analysis of the median-plane HRTFs in the CIPIC HRTF database. It was verified that the 12 PCs can be general basis functions to model arbitrary subjectpsilas median-plane HRIRs, which are not included in the process to obtain the basis functions, through the quantitative analysis on the modeling error in the least-squares sense and the subjective listening tests. A HRTF customization method based on subjective tuning of the general basis functions was proposed. In the subjective listening test results, all subjects reported better performances for the vertical perception and the front-back discrimination with the customized HRTFs than those with the non-individualized HRTFs except for the front-back confusions of one subject.

Median HRIR Customization via Principal Components Analysis

A principal components analysis of the entire median HRIRs in the CIPIC HRTF database reveals that the individual HRIRs can be adequately reconstructed by a linear combination of several orthonormal basis functions. The basis functions represent the inter-individual and inter-elevation variations in median HRIRs. There exist elevation-dependent tendencies in the weights of basis functions, and the basis functions can be ordered according to the magnitude of standard deviation of the weights at each elevation. We propose a HRIR customization method via tuning of the weights of 3 dominant basis functions corresponding to the 3 largest standard deviations at each elevation. Subjective listening test results show that both front-back reversal and vertical perception can be improved with the customized HRIRs.