Multilingual speaker recognition using ANFIS

Abstract

Feature based Recognition Systems has been an area of intense research for long. The creation of a reliable, robust and sufficiently efficient recognition system has been tried using features from several sources including textual and image sources. Speech based sources have also been used for the creation of such a recognition system. However, variations caused due to differences in individual speaker characteristics, mood variations and inter-mingled noise disturbances make the realization of such a system very difficult. This paper proposes a recognition system for identification of the speaker, language and the words spoken. The system makes use of Adaptive Neuro-Fuzzy Inference paradigm for the same. First, the sampling frequency and the speech features are extracted from the speech database to form speech feature vectors. The features used are LPC, LPCC, RC, LAR, LSF and ARSCIN. The speech database is prepared using 25 speakers including male and female speakers. Five different speaking texts of different languages having same meaning are used to get the best speaker identification accuracy. The languages spoken by the speakers include English, Hindi, Punjabi, Sanskrit and Telugu. The Feature vectors, thus prepared, are fed to an Adaptive Neuro-Fuzzy Inference System for speaker, language and word recognition. The experimental results show the system to be amply efficient and successful in the recognition tasks involved.

An adaptive neuro-fuzzy inference system for the qualitative study of perceptual prominence in linguistics

“…In Speech and Natural Language Processing, ANFIS has been already proposed, for instance, to model the relationship between acoustic features and emotion dimension [21], to predict the imprecise nature of speech prosody [22] and to identify the speaker, language and the words spoken [23]. However, no work exists about the application of ANFIS for a prominence study.…”

Adaptive Network Based Fuzzy Inference System For Speech Recognition Through Subtractive Clustering

“…In [18] ANFIS was performed to reduce noise and enhance speech, also for the recognition of isolated digits with speaker-independent [19]. Speaker, language and word recognition was completed by ANFIS [20], furthermore for caller behavior classification [21]. All of these works had used clustering techniques to determine the structure of ANFIS.…”

Combination of Features for Multilingual Speaker Identification with the Constraint of Limited Data

“…The textdependent identification requires the speaker to produce speech for the same text, both during training and testing whereas the text-independent identification does not rely on a specific text being spoken [3]. Text-independent speaker identification systems are more versatile, but their accuracy is considerably lower than that of text-dependent systems [4]. To achieve acceptable results in this case, more speech data is usually necessary for both training and testing purposes [4].…”

“…They simulated a model of ANN based multilingual speaker recognition system for eight Indian languages (Hindi, English, Assami, Telugu, Punjabi, Rajasthani, Marathi and Bengali) and achieved a success rate of 95%. Bipul Pandey et al [4] proposed a multilingual speaker recognition scheme using adaptive neuro fuzzy inference scheme (ANFIS) for the identification of the speaker and the words spoken. A robust and sufficiently efficient recognition system was tried by them using the features obtained from several sources including the textual and image sources, which produced excellent results.…”

“…Text-independent speaker identification systems are more versatile, but their accuracy is considerably lower than that of text-dependent systems [4]. To achieve acceptable results in this case, more speech data is usually necessary for both training and testing purposes [4]. Speaker recognition in limited data condition aims at recognizing speaker with the constraint that both training and testing data are limited [5].…”