Showing papers by "Goutam Saha published in 2006"
01 Sep 2006
TL;DR: A novel method of boundary estimation algorithm that extensively utilizes biomedical domain features for reduction of time and computational complexities and is more accurate is proposed.
Abstract: Determining the exact timings of the cardiac events by the first (S1) and second (S2) heart sounds from the phonocardiogram (PCG) signal, represents a great challenge, specially in pathological cases. A system that allows the detection of heart diseases, is needed a properly boundary estimated cardiac cycles of heart sound signal. Boundary estimation algorithm gives the exact boundary of the S1 and S2 of PCG signal. This paper proposes a novel method of boundary estimation algorithm that extensively utilizes biomedical domain features for reduction of time and computational complexities and is more accurate. The performance of the algorithm is evaluated for 10 commonly occurring pathological cases and normal heart sound for various sampling frequencies, recording environments and age group of subjects. It is found to give an accuracy of 96.30 percent for proposed boundary estimation algorithm.
14 citations
01 Sep 2006
TL;DR: The work introduces the fusion of log Gabor wavelet and maximum a posteriori (MAP) estimator for robust text-independent SI system and the Gaussian mixture model (GMM) is shown to outperform the other methods.
Abstract: Speaker Identification (SI) system needs an efficient feature extraction process and an appropriate speaker model developed from these features. The work introduces the fusion of Log Gabor Wavelet (LGW) and Maximum a Posteriori (MAP) Estimator for robust Text-independent SI system. The focus of this paper is on the robustness to degradations produced by transmission over a telephone channel. Complete experimental framework is conducted on 49 speakers, conversational telephone King-92 SI speech database with two well known speaker models i.e. Gaussian mixture model (GMM) and Vector Quantization (VQ). Comparisons are made with two different established methods as well as with normal feature extraction procedure to show the robustness of the new approach in different time segments. The GMM attains 98.8% of identification accuracy using 30 second of wide band speech utterances and 87.3% of identification accuracy using 30 second of narrow band speech utterances and is shown to outperform the other methods.