Browsing by Author "Lin, Tse-En"

Now showing 1 - 3 of 3

Spectro-temporal Modulation Based Singing Detection Combined with Pitch based Grouping for Singing Voice Separation
(2013-01-01)
Lin, Tse-En
;
Hsu, Chung-Chien
;
Chen, Yi-Cheng
;
Chen, Jian-Hueng
;
Chi, Tai-Shih
A spectro-temporal modulation based singing voice detection cascaded with a Viterbi based pitch tracking algorithm is proposed in this paper for singing-voice separation from monaural recordings. To detect the singing voice, the spectrotemporal modulation energy related to voice harmonics is extracted using a spectro-temporal modulation analysis framework developed for the Fourier spectrogram. Separation of singing -voice from background music is conducted using a binary mask to group estimated harmonics of singing voice. The proposed system is evaluated using MIR-1K dataset and is shown outperforming three other binary-mask based systems in the vocal/music separation task.
SPECTRO-TEMPORAL SUBBAND WIENER FILTER FOR SPEECH ENHANCEMENT
(2012)
Hsu, Chung-Chien
;
Lin, Tse-En
;
Chen, Jian-Hueng
;
Chi, Tai-Shih
;
電子工程學系及電子研究所
In this paper, we propose a signal-channel speech enhancement algorithm by applying the conventional Wiener filter in the spectro-temporal modulation domain. The multi-resolution spectro-temporal analysis and synthesis framework for Fourier spectrograms [12] is extended to the analysis-modification-synthesis (AMS) framework for speech enhancement. Compared with conventional speech enhancement algorithms, a Wiener filter and an extended minimum mean-square error (MMSE) algorithm, our proposed method outperforms them by a large/small margin in white/babble noise conditions from both objective and subjective evaluations.
VOICE ACTIVITY DETECTION BASED ON FREQUENCY MODULATION OF HARMONICS
(2013)
Hsu, Chung-Chien
;
Lin, Tse-En
;
Chen, Jian-Hueng
;
Chi, Tai-Shih
;
電機資訊學士班
In this paper, we propose a voice activity detection (VAD) algorithm based on spectro-temporal modulation structures of input sounds. A multi-resolution spectro-temporal analysis framework is used to inspect prominent speech structures. By comparing with an adaptive threshold, the proposed VAD distinguishes speech from non-speech based on the energy of the frequency modulation of harmonics. Compared with three standard VADs, ITU-T G. 729B, ETSI AMR1 and AMR2, our proposed VAD significantly outperforms them in non-stationary noises in terms of the receiver operating characteristic (ROC) curves and the recognition rates from a practical distributed speech recognition (DSR) system.