Abstract

Human-computer interaction plays a significant role for literate/illiterate and visually challenged users to access information. The mode of human-computer interaction could be speech, text, gestures, facial expressions, symbols etc., or a combination of these. An interaction could be in the form of a conversation including statements, questions, answers and expressions. A (uni-modal or multi-modal) conversational type of human-computer interaction is often referred to as a conversational system. A conversational system with speech as an input mode assumes significance as speech is a natural means of communication for human beings. The goal of a speech-based conversation (SBC) system is to provide information by conversing with a human-being in a natural fashion. Our objective is to develop speech based conversational systems for in formation access in an Indian language.

Abstract

Glottal closure instants (GCI) information is useful for accurate speech analysis. In particular accurate spectrum analysis is performed by considering the speech in the intervals of glottal closure. In this paper we propose an approach for detection of GCI based on Bessel features, amplitude and frequency modulated (AM-FM) signal. Using appropriate range of Bessel coefficients, the narrow band, band limited signal is obtained for the given signal. The band limited signal is considered as a AM-FM signal. The signal is band limited for0-300 Hz to remove effect of formants. Amplitude Envelope (AE)function of the AM-FM signal model has been estimated by the discrete energy separation algorithm (DESA). We experimentally evaluated our approach to detect GCI on CMU-Arctic database.The corresponding electro-glottograph (EGG) signals are used as a reference for the validation of the detected GCI locations.

Abstract

Voice onset time is an important temporal feature which is often overlooked in speech perception, speech recognition as well as accent detection. The VOT in unvoiced stops varies with a number of factors, among which the most established one is the place of articulation. In this paper we propose an approach for the automatic detection of VOT. The proposed method uses Bessel expansion to emphasize the vowel and consonant regions of stop consonant vowel units (SCV) such as /ka/, /Ta/, /ta/ and /pa/. AM-FM signal has been emphasized after appropriate consideration of the range of Bessel coefficients,separately for the vowel and consonant regions of SCV units. There constructed signal from the Bessel expansion is a narrow-band AM-FM signal, therefore the amplitude envelope (AE) function for the emphasized signal can be estimated using discrete energy separation algorithm (DESA). For the detection of VOT, both the AE of vowel and consonat emphasized signal has been analyzed.Detection of VOT is analyzed for the continuous speech corpus consisting of recording television broadcast news bulletins.

Abstract

In this paper, we explore Bessel features to determine the number of speakers from multispeaker speech signals collected simultaneously from a pair of spatially separated microphones. The arrival of the speech signals from speaker to microphones gives the time delays of given speaker. The time delays can be estimated by performing the cross-correlation to the band limited multispeaker signals collected at the two microphones. Signals are band limited using a finite number of Bessel coefficients. The computer simulation results demonstrate the proposed method is efficient compared to existing methods.

Abstract

The Independent Component Analysis with Reference (ICA - R) also called as constrained ICA (cICA) extracts only the desired source signals from the mixture of source signals by incorporating some prior information into the separation process. To overcome the problem of designing the reference signal when there is no prior information about the desired signal in the cICA, an improved method is proposed to use a different speech signal generated by the same physical source. The cICA is extended to use Bessel coefficients of the observed signals and the reference signal for processing as they converge faster than the other transformations. Since the Bessel functions provide the desired properties, efficient in representing speech signals, less memory storage they have been exploited in speech processing [1]. The results demonstrate the efficiency of the proposed method

Abstract

In many circumstances in image processing, image resizing is done, either to magnify or to reduce the size of a digital image. The spatial domain based resizing methods such as bilinear interpolation and bi-cubic interpolation are simple and work better for image size magnification. The main drawback in using them is that they are not suitable for image size reduction. In this paper, we propose a new method for image resizing based on Bessel transform (BT). The performance of image resizing based on BT is compared to that of spatial domain based resizing techniques, the results are viewed in terms of Peak Signal to Noise ratio (PSNR) and Mean Square Error (MSE). Experimental results confirm that the proposed method maintains better image quality when image size is enlarged and also when image size is reduced.

Abstract

Speech analysis is traditionally performed using short-time analysis to extract features in time and frequency domains. The window size for the analysis is fixed somewhat arbitrarily, mainly to account for the time varying vocal tract system during production. However, speech in its primary mode of excitation is produced due to impulse-like excitation in each glottal cycle. Anchoring the speech analysis around the glottal closure instants (epochs) yields significant benefits for speech analysis.Epoch-based analysis of speech helps not only to segment the speech signals based on speech production characteristics, but also helps in accurate analysis of speech.It enables extraction of important acoustic-phonetic features such as glottal vibra-tions, formants, instantaneous fundamental frequency, etc. Epoch sequence is useful to manipulate prosody in speech synthesis applications. Accurate estimation of epochshelps in characterizing voice quality features. Epoch extraction also helps in speech enhancement and multi speaker separation. In this tutorial article, the importance of epochs for speech analysis is discussed, and methods to extract the epoch informa-tion are reviewed. Applications of epoch extraction for some speech applications are demonstrated.

Abstract

This paper describes the challenges faced in language learning and use, with a focus on domain-specific vocabulary learning. We focus on the architecture and the presentation model of domain-specific knowledge components. The paper first discusses the significance of assistive mechanisms for vocabulary learning in an academic domain and then outlines two architectures for the presentation model. We suggest how the design of this environment for vocabulary use and learning in the academic domain must adopt different presentations and search strategies to confront the specific challenges in the domain. As this problem can only be solved on a long term basis, we also briefly describe a set of games as a first step in the implementation of the assistive mechanisms.

Abstract

In this paper we describe the development of Mandi Information System, a Telugu spoken dialogue system for obtaining price information of agricultural commodities like vegetables, fruits, pulses, spices, etc.. The target users of MIS are primarily the farmers in rural and semiurban areas. Speech recognition is error prone and it is necessary for the dialogue system to make minimum number of errors while acquiring information from a user and also to detect errors (if not correctable) and adopt appropriate strategies. In this paper we suggest an approach to improve the performance and usability of the system by using multiple decoders and contextual information

Abstract

This paper proposes the Fourier-Bessel cepstral coefficients (FBCC) as features for robust text-independent speaker identification. Fourier-Bessel (FB) expansion is used instead of Fourier transform for representing the signal in frequency domain. FB expansion can be viewed as two-dimensional Fourier transform. Change in the kernel of the transform from exponential to decaying exponentials helps in viewing the speech signal as a linear sum of decaying exponentials. For signals arising out of acoustic tubes, where the signal is subjected to many damping effects, delays in the different components of the signal is inevitable. Representing such signals using FB coefficients helps in able identification of different components present in the signal. The random non-stationary nature of speech signal is more efficiently represented by damped sinusoidal nature of basis function that is more natural for the voiced speech signal since Bessel functions have damped sinusoidal as basis function, so it is more natural choice for the representation of natural signals. Vocal tract is modeled as a set of linear acoustic tubes being cylindrical in shape can be efficiently modeled using FB expansion because Bessel functions are solutions to cylindrical wave equations. The proposed approach to speaker identification is based on FBCC features, and method employ Gaussian mixture for modeling the speaker characteristics. However, we have build the speaker models from the Fourier-Bessel features derived from the speech samples, as an alternative to Mel-frequency cepstral coefficients (MFCC) and linear prediction cepstral coefficients (LPCC) for building the speaker models. An evaluation of the Gaussian mixture model is conducted on TIMIT database which consists of 630 speakers and 10 speech utterances per speaker and white noise signals of TIMIT database having various SNRs of 50, 40, 30 and 20 dB. Using the statistical model like Gaussian mixture model (GMM) and features extracted from the speech signals build a unique identity for each person who enrolled for speaker identification [1]. Estimation and Maximization algorithm is used for finding the maximum likelihood solution for a model with features, to test the later speeches against the database of all speakers who enrolled in the database. Experimental results shows that the FBCC can be used as the alternate feature for the LPCC and MFCC since it can improve the performance of the speaker identification task