Search Results 1-20 of 54

  • 1 / 3
  • Stereophonic Music Separation Based on Non-Negative Tensor Factorization with Cepstral Distance Regularization

    SEKI Shogo , TODA Tomoki , TAKEDA Kazuya

    … To address this issue, this paper proposes a Cepstral Distance Regularization (CDR) method for NTF-based stereo channel separation, which involves making the cepstrum of the separated source signals follow Gaussian Mixture Models (GMMs) of the corresponding the music source signal. … Experimental results demonstrate that the proposed method yields significant improvements within both separation frameworks, and that cepstral distance regularization provides better separation parameters.</p> …

    IEICE Transactions on Fundamentals of Electronics, Communications and Computer Sciences E101.A(7), 1057-1064, 2018

    J-STAGE

  • A method for estimating vocal-tract shape from a target speech spectrum

    Kaburagi Tokihiko

    … Our method effectively optimizes the vocal-tract shape when combined with the perturbation relationship between the speech spectrum parameters (i.e., cepstral parameters) and formants. …

    Acoustical Science and Technology 36(5), 428-437, 2015

    J-STAGE

  • A study for estimating the vocal-tract shape from speech spectrum using a sensitivity function  [in Japanese]

    KABURAGI Tokihiko

    … To determine vocal-tract shape from speech, this paper presents an inversion method for simultaneously estimating the cross-sectional area and length of the vocal tract from cepstral parameters. … When combined with a perturbation relationship of speech spectrum parameters (i.e., cepstral parameters) and formants, our method effectively optimizes the vocal-tract shape. …

    IEICE technical report. Speech 114(303), 47-52, 2014-11-13

  • Integration of Spectral Feature Extraction and Modeling for HMM-Based Speech Synthesis

    NAKAMURA Kazuhiro , HASHIMOTO Kei , NANKAKU Yoshihiko , TOKUDA Keiichi

    … In the feature extraction module, the statistical mel-cepstral analysis technique has been used and the objective function is the likelihood of mel-cepstral coefficients for given speech waveforms. … In the acoustic modeling module, the objective function is the likelihood of model parameters for given mel-cepstral coefficients. …

    IEICE Transactions on Information and Systems E97.D(6), 1438-1448, 2014

    IR J-STAGE

  • A Study on Speaker Recognition Based on Decomposition of Periodic and Aperiodic Components  [in Japanese]

    ISHIKAWA Yuki , NISHIDA Masafumi , FUJIMOTO Masakiyo , YAMAMOTO Seiichi

    … In conventional researches, mel-frequency cepstral coefficients (MFCC) are widely used for a feature parameter which approximates vocal tract characteristics. … However, use of other feature parameters, i.e., a fluctuation of the speech signal has not been fully discussed in recent research progress. …

    IEICE technical report. Speech 112(422), 25-30, 2013-01-30

  • Cascaded Subband Energy-Based Emotion Classification

    AMARAKEERTHI Senaka , MORIKAWA Chamin , NWE Tin Lay , DE SILVA Liyanage C. , COHEN Michael

    … Speech is modulated for different emotions by varying frequency- and energy-related acoustic parameters such as pitch, energy, and formants. … In this research, Two-Layered Cascaded Subband Cepstral Coefficients (TLCS-CC) analysis was introduced to study energy variations within low and high arousal emotions as a novel approach for emotion classification. … The new approach was compared with Mel frequency cepstral coefficients (MFCC) and log frequency power coefficients (LFPC). …

    IEEJ Transactions on Electronics, Information and Systems 133(1), 200-210, 2013-01-01

    J-STAGE References (33)

  • Blind Restoration of Motion Blurred Image by Applying a Non-iterative Algorithm

    Kuroyanagi Shinichi , Maruo Ryota , Kubo Yukihiro , Sugimoto Sueo

    … We explain the motion degradation analysis and how to estimate parameters using Cepstral analysis, Radon transform and Hough transform. …

    Proceedings of the ISCIE International Symposium on Stochastic Systems Theory and its Applications 2013(0), 94-100, 2013

    J-STAGE

  • Speaker Identification Using Japanese Monosyllables and Contributions of Nasal Consonants and Vowels to Identification Accuracy

    Amino Kanae , Osanai Takashi

    … Nasal consonants and the following vowels were excerpted from the recorded monosyllables, and 30th-order cepstral coefficients were calculated for each as acoustic features. … Analysis was performed on frequently confused speaker pairs using 11 parameters for vowels and six parameters for nasals. … The parameters were selected from various attributes related to the physiological properties of the speakers and the acoustic properties of their speech. …

    Japanese Journal of Forensic Science and Technology 18(1), 13-21, 2013

    J-STAGE

  • Intra-speaker spectral parameter variation between utterances of the same sentence and its prediction  [in Japanese]

    INUKAI Tatsuo , TODA Tomoki , NEUBIG Graham , SAKTI Sakriani , NAKAMURA Satoshi

    … In spectral conversion of statistical voice conversion technologies, distance measures between the converted and target parameters, such as mel-cepstral distortion, are often used as evaluation/training metrics. … However, even if the same speaker utters the same sentence, the spectral parameters of those utterances vary, and therefore, a distance between them still exists. …

    IEICE technical report. Speech 112(281), 13-18, 2012-11-01

    References (11)

  • Human Activity Recognition from Environmental Background Sounds for Wireless Sensor Networks

    ZHAN Yi , NISHIMURA Jun , KURODA Tadahiro

    … Sound feature extraction Mel Frequency Cepstral Coefficients (MFCC) and Vector Quantization (VQ) classification Linde-Buzo-Gray algorithm (LBG) algorithms are applied for recognizing the background sounds in the human daily activities. … In these algorithms, how three parameters (i.e., Mel filters number, frame-to-frame overlap and LBG codebook cluster number) affect system's calculation burden and accuracy is also investigated. …

    IEEJ Transactions on Electronics, Information and Systems 130(4), 565-572, 2010-04-01

    J-STAGE References (23)

  • The measurement of the detecting the sound signals in the human body for prediction of sleep in sleep-wake state (The 2nd report)  [in Japanese]

    AKITA Masanori , MIDORIKAWA Yoichi , KANECHIKA Tatsuya , SAKAGUCHI Masakazu

    … However, the systems using these parameters are fairly large and complicated. … The signals are analyzed into the cepstral coefficients and the feature parameters are defined using these coefficients. … The feature parameters are compared with the signals obtained from the piezoelectric sensors, which is effective to detect the sleep in sleep-wake state. …

    IEICE technical report 109(286), 25-30, 2009-11-12

    References (12)

  • An MTF-based method for blindly restoring bone-conducted speech  [in Japanese]

    KINUGASA Kota , UNOKI Masashi , AKAGI Masato

    … We propose a method of determining the parameters of the MTF-based model without AC speech. … The distances based on LP coefficients and mel-frequency cepstral coefficients were used to evaluate the improvements in cepstral distances that are useful for ASR systems. …

    IEICE technical report 109(100), 79-84, 2009-06-18

    References (14)

  • Multi-Input Feature Combination in the Cepstral Domain for Practical Speech Recognition Systems

    OBUCHI Yasunari , HATAOKA Nobuo

    … In this paper we describe a new framework of feature combination in the cepstral domain for multi-input robust speech recognition. … The general framework of working in the cepstral domain has various advantages over working in the time or hypothesis domain. … Finally, we propose to take another advantage of working in the cepstral domain. … We also refer to automatic optimization of some parameters in the proposed algorithms. …

    IEICE Transactions on Information and Systems 92(4), 662-670, 2009-04-01

    J-STAGE References (23)

  • Human Activity Recognition Based on Acceleration Information  [in Japanese]

    TAKEUCHI Shinichi , ITOU Shinya , TAMURA Satoshi , HAYAMIZU Satoru

    … As the feature parameters, we compared time series of accelerometer signal and frequency domain parameters whose are obtained by speech signal processing. … We use MFCC (Mel-Frequency Cepstral Coefficients) which emphasizes low frequency, angular information. …

    IEICE technical report 108(453), 229-234, 2009-02-23

    References (10) Cited by (1)

  • Human Activity Recognition Based on Acceleration Information  [in Japanese]

    TAKEUCHI Shinichi , ITOU Shinya , TAMURA Satoshi , HAYAMIZU Satoru

    … As the feature parameters, we compared time series of accelerometer signal and frequency domain parameters whose are obtained by speech signal processing. … We use MFCC (Mel-Frequency Cepstral Coefficients) which emphasizes low frequency, angular information. …

    IEICE technical report 108(454), 229-234, 2009-02-23

    References (10)

  • Human Activity Recognition Based on Acceleration Information  [in Japanese]

    TAKEUCHI Shinichi , ITOU Shinya , TAMURA Satoshi , HAYAMIZU Satoru

    … As the feature parameters, we compared time series of accelerometer signal and frequency domain parameters whose are obtained by speech signal processing. … We use MFCC (Mel-Frequency Cepstral Coefficients) which emphasizes low frequency, angular information. …

    IEICE technical report 108(455), 229-234, 2009-02-23

    References (10)

  • The examination of the detecting the sound signals in the human body for prediction of sleep in sleep-wake state  [in Japanese]

    AKITA Masanori , MIDORIKAWA Yoichi

    … However, the systems using these parameters are fairly large and complicated. … The signals are analyzed into the cepstral parameters and the feature parameters are defined using these coefficients. … The feature parameters are compared with the signals obtained from the piezoelectric sensors, which is effective to detect the sleep in sleep-wake state. …

    IEICE technical report 108(306), 93-98, 2008-11-13

    References (10) Cited by (1)

  • Unsupervised Phoneme Segmentation Using Mahalanobis Distance

    QIAO Yu , MINEMATSU Nobuaki

    … An objective function, summation of squared error (SSE), is developed by using Euclidean distance of cepstral features. … The essential problem here is how to determine the parameters (covariance matrix) for MD calculation. … We deal with this problem in a learning framework and propose two criteria for determining the optimal parameters: Minimum of Summation Variance (MSV) and Maximum of Discrimination Variance (MDV). …

    IEICE technical report 107(551), 69-74, 2008-03-13

    References (12)

  • Evaluation of a Noise-Robust Multi-Stream Speaker Verification Method Using F_0 Information

    ASAMI Taichi , IWANO Koji , FURUI Sadaoki

    … and cepstral features. … features improve the verification performance in various noisy environments, and that our stream-weight and threshold optimization method effectively estimates control parameters so that FARs and FRRs are adjusted to achieve equal error rates (EERs) under various noisy conditions. …

    IEICE Transactions on Information and Systems 91(3), 549-557, 2008-03-01

    IR J-STAGE References (23) Cited by (1)

  • A New Approach of Blind Image Deconvolution Method without Artifact along the Blurring Paths

    Maki W. F. Al , Shimahashi T. , Kitagawa T. , Sugimoto S.

    … Point spread function (PSF) of the linear motion blur has two parameters, i.e., motion length and motion direction. … The parameters are estimated using the modified discrete Radon transform and cepstral analysis. … The estimated PSF parameters are then used in the image deconvolution process. …

    Proceedings of the ISCIE International Symposium on Stochastic Systems Theory and its Applications 2008(0), 195-200, 2008

    J-STAGE

  • 1 / 3
Page Top