Continuous Bangla Speech Segmentation using Short-term Speech Features Extraction Approaches

Abstract

This paper presents simple and novel feature extraction approaches for segmenting continuous Bangla speech sentences into words/sub-words. These methods are based on two simple speech features, namely the time-domain features and the frequency-domain features. The time-domain features, such as short-time signal energy, short-time average zero crossing rate and the frequency-domain features, such as spectral centroid and spectral flux features are extracted in this research work. After the feature sequences are extracted, a simple dynamic thresholding criterion is applied in order to detect the word boundaries and label the entire speech sentence into a sequence of words/sub-words. All the algorithms used in this research are implemented in Matlab and the implemented automatic speech segmentation system achieved segmentation accuracy of 96%.

Authors and Affiliations

Md Mijanur Rahman , Md. Al-Amin Bhuiyan

Keywords

Related Articles

Automatic Fall Detection using Smartphone Acceleration Sensor

In this paper, we describe our work on developing an automatic fall detection technique using smart phone. Fall is detected based on analyzing acceleration patterns generated during various activities. An additional long...

A Defeasible Logic-based Framework for Contextualizing Deployed Applications

In human to human communication, context increases the ability to convey ideas. However, in human to application and application to application communication, this property is difficult to attain. Context-awareness becom...

Access Control Model for Modern Virtual e-Government Services: Saudi Arabian Case Study

e-Government services require intensive information exchange and interconnection among governmental agencies to provide specialized online services and allow informed decision-making. This could compromise the integrity,...

Reliability and Connectivity Analysis of Vehicluar Ad Hoc Networks for a Highway Tunnel

Vehicular ad-hoc network (VANET) uses ‘mobile internet’ to facilitate the communication between vehicles and with the goal to ensure road safety and achieve secure communication. Thus the reliability of this type of netw...

Task Scheduling Frameworks for Heterogeneous Computing Toward Exascale

The race for Exascale Computing has naturally led computer architecture to transit from the multicore era and into the heterogeneous era. Many systems are shipped with integrated CPUs and graphics processing units (GPUs)...

Download PDF file
  • EP ID EP135420
  • DOI 10.14569/IJACSA.2012.031121
  • Views 113
  • Downloads 0

How To Cite

Md Mijanur Rahman, Md. Al-Amin Bhuiyan (2012). Continuous Bangla Speech Segmentation using Short-term Speech Features Extraction Approaches. International Journal of Advanced Computer Science & Applications, 3(11), 131-138. https://www.europub.co.uk/articles/-A-135420