This research paper aims to make essential developments in Speech Recognition(SR), the compiler gives the user a choice to choose the type of output, whether it is textual or conversational (audio). Many large companies have developed such Speech Recognition Systems (SRS), especially the companies producing Smartphones, Computers, and Laptops. If translation is taken as a model application, they have not yet developed the perfect systems. The purpose of this paper is to add facilities to the Speech Recognition (SR) software so that it can deal with spoken languages.

Implementation of a Verbal Compiler: The Need to Develop Audio Language to Keep Pace with Rapid Development becomes a Necessity

This paper introduces a new back-end classifier for a speech recognition system that is based on artificial life (ALife) The ALife species being used for classification purposes are called wains, which were developed using the Creatur framework The speech recognition task used in the evaluation of the new classifier is that of isolated digit recognition Performance of the proposed back-end classifier is benchmarked for comparison purposes against a hidden Markov model (HMM)-based classifier Mel frequency cepstral coefficients extracted from the speech by a front-end processor are used by the two back-ends for classification of the speech utterances In tests carried out to date, the performance of the proposed classifier falls short of the performance of the HMM-based classifier However, the results suggest that with further training the recognition accuracy of the proposed classifier can be significantly improved

Evaluation of wains as a classifier for automatic speech recognition

We present the Self-Generating Model (SGM), a new version of the Self-organizing Map (SOM) that has been adapted for use in intelligent data mining ALife agents. The SGM sacrifices the topology-preserving ability of the SOM, but is equally accurate, and faster, at identifying handwritten numerals. It achieves a higher accuracy faster than the SOM. Furthermore, it increases model stability and reduces the problem of “wasted” models. We feel that the SGM could be a useful alternative to the SOM when topology preservation is not required.

The Self-Generating Model: An Adaptation of the Self-organizing Map for Intelligent Agents and Data Mining

Speech recognition is an important technology and can be used as a great aid for individuals with sight or hearing disabilities today. There are extensive research interest and development in this area for over the past decades. However, the prospect in Malaysia regarding the usage and exposure is still immature even though there is demand from the medical and healthcare sector. The aim of this research is to assess the quality and the impact of using computerized method for early screening of speech articulation disorder among Malaysian such as the omission, substitution, addition and distortion in their speech. In this study, the statistical probabilistic approach using Hidden Markov Model (HMM) has been adopted with newly designed Malay corpus for articulation disorder case following the SAMPA and IPA guidelines. Improvement is made at the front-end processing for feature vector selection by applying the silence region calibration algorithm for start and end point detection. The classifier had also been modified significantly by incorporating Viterbi search with Genetic Algorithm (GA) to obtain high accuracy in recognition result and for lexical unit classification. The results were evaluated by following National Institute of Standards and Technology (NIST) benchmarking. Based on the test, it shows that the recognition accuracy has been improved by 30% to 40% using Genetic Algorithm technique compared with conventional technique. A new corpus had been built with verification and justification from the medical expert in this study. In conclusion, computerized method for early screening can ease human effort in tackling speech disorders and the proposed Genetic Algorithm technique has been proven to improve the recognition performance in terms of search and classification task.

Malay articulation system for early screening diagnostic using hidden markov model and genetic algorithm

Speech has always been the most important role in our day to day communication. People used to represent our thoughts through the some known language, which may be done through human to human interaction or human to machine interaction. Human Machine Interaction (HMI) plays a vital role on interaction in speech recognition system. Speech or word by word recognition process is finishing the process of separating the speech characteristics from the pre-recorded datasets, and classifying the same characteristics. A word or speech must be forwarded to more advanced software for syntactic and semantic analysis in order to be recognized. So the speech recognition system has to find a very good feature extraction method, speech classifier and performance evaluator to recognize well in all noisy environments. Hence this paper provides the overview of challenges accepted by different researchers as well as it provides the collective information about different assessment, methodology like Hidden Markov Model and Gaussian Mixture Model with MFCC and LSTM and performance improvement given by the researchers for accepted challenges to do the speech recognition in noisy environment.

A Review on Quality Speech Recognition Under Noisy Environment

After years of research activity, the machine recognition performance of speech still does not match human performance. As speech recognition is a pattern recognition problem, classification is an important part of any speech recognition system. This paper, briefly reviews a number of classification methods that have been used in automatic speech recognition systems and proposes a new back-end classifier that is based on artificial life. The paper describes how the proposed classifier can be used in a speech recognition system.

Automatic speech recognition using artificial life

Research in speech recognition has produced different approaches that have been used for the classification of speech utterances in the back-end of an automatic speech recognition (ASR) system. As speech recognition is a pattern recognition problem, classification is an important part of any speech recognition system. This paper proposes a new back-end classifier that is based on artificial life (ALife) and describes how the proposed classifier can be used in a speech recognition system.

Rosemary T. Salaja

Papers

Automatic speech recognition using artificial life

A Life-Based Classifier for Automatic Speech Recognition

Evaluation of wains as a classifier for automatic speech recognition