Feature selection algorithms for automatic speech recognition

Kalamani, M.; Valarmathy, S.; Poonkuzhali, C.; Catherine, J N

doi:10.1109/iccci.2014.6921797

Cited by 9 publications

(3 citation statements)

References 16 publications

(18 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…There is several segmentation approaches used since the past decade. Segmentation approach refers to the method of decomposing the speech signal into the fraction of basic phonetic units [10]. Speech signal can be segmented on different level: phonemic, sub phonemic, syllabic, word level, syntagmatic level depending on the segmentation algorithm used.…”

Section: Introductionmentioning

confidence: 99%

Emotional speech feature selection using end-part segmented energy feature

Zaidan

Salam

2019

IJEECS

View full text Add to dashboard Cite

The accuracy of human emotional detection is crucial in the industry to ensure effective conversations and messages delivery. The process involved in identifying emotions must be carried out properly and using a method that guarantees high level of emotional recognition. Energy feature is said to be a prosodic information encoder and there are still studies on energy use in speech prosody and it motivate us to run an experiment on energy features. We have conducted two sets of studies: 1) whether local or global features that contribute most to emotional recognition and 2) the effect of the end-part segment length towards emotion recognition accuracy using 2 types of segmentation approach. This paper discussed about Absolute Time Intervals at Relative Positions (ATIR) segmentation approach and global ATIR (GATIR) using end-part segmented global energy feature extracted from Berlin Emotional Speech Database (EMO-DB). We observed that global feature contribute more to the emotional recognition and global features that are derived from longer segments give higher recognition accuracy than global feature derived from short segments. The addition of utterance-based feature (GTI) to ATIR segmentation somewhat contributes to increase the accuracy by 5% up to 8% and conclude that GATIR outperformed ATIR segmentation approached in term of its higher recognition rate. The results of this study where almost all the sub-tests provide an increased result proving that global feature derived from longer segment lengths acquire more emotional information and enhance the system performance.

show abstract

Section: Introductionmentioning

confidence: 99%

Emotional speech feature selection using end-part segmented energy feature

Zaidan

Salam

2019

IJEECS

View full text Add to dashboard Cite

show abstract

“…Speech Recognition known as Automatic Speech Recognition (ASR) is the procedure of converting speech signals to a sequence of words through methods of algorithms accomplished as a computer programme [1]. Automatic speech recognition enables the computer to identify words that are spoken into microphone or a telephone.…”

Section: Introductionmentioning

confidence: 99%

Voice Biometrical Match of Mother and Daughter for Forensic Purpose

Kaur¹,

Verma²

2019

IJFP

View full text Add to dashboard Cite

The similarity in voices among blood relations has been always a fascinating subject in forensic phonetics. The present work is an initiatory study of investigational character that explains the resemblance of mother and her daughter phonation. This study is an attempt to report a novel approach to the speaker recognition and comparison of the voice of mother and her daughter by the use of voice biometric software. An attempt has been made to enlighten the role of voice as evidence in forensic investigation. Having noted a lack of oral corpora of mother and daughter voices, we have collected a database which consists of the 40 samples in a combination 20 each of mother & daughter respectively. Voice transcripts of mother and daughter were analyzed & compared using speech analysis software to examine the similarities & variations in sound spectrograms. The results of our study showed that the mother and daughter voice samples have higher similarities which were estimated to 88%-92%. It is due to the influence of genetic endowment, physical similarities and inheritance of vocal gesture & posture. Voice Biometrical Match cannot be 100% due to law of individuality. This suggests that the system features are to a great extent genetically conditioned and that they are hence useful and robust for comparing speech samples of known and unknown origin, as found in legal cases. This research shall enlighten an idea of getting corroborative evidence, further which can be a part of Forensic Investigation. Forensic experts can contribute to society/law with advance as well as fast crime detection solutions challenge in speaker identification.

show abstract

“…Therefore, feature selection can be employed to eliminate the redundant and irrelevant features present in a high dimensional feature set. Traditionally, a single feature subset is obtained for distinguishing each class from the other classes [1]. On the contrary, class-specific feature sets consisting of different feature subsets for each class can also be obtained [2].…”

Section: Introductionmentioning

confidence: 99%

Automatic Speech Segmentation and Recognition using Class-Specific Features

Rekha¹,

Chatrapati²,

Babu³

2015

IJCA

View full text Add to dashboard Cite

The class-specific automatic speech recognition systems construct an individual classifier for each class based on its own feature set, wherein the feature set for each class is selected such that it distinguishes that class from the other classes most accurately. Consequently, different feature set sequences must be fed into each of the classifiers, and the output of each of the classifiers must be combined to predict the actual class of the observation sequences. However, speech is continuous, and to be able to apply class-specific features, speech should be segmented and fed to the classifiers, which requires the identification of segmentation cues. This paper proposes a framework that jointly segments, and combines the output of the class-specific classifiers in the absence of any segmentation cues using a recursive formulation.

show abstract

Feature selection algorithms for automatic speech recognition

Cited by 9 publications

References 16 publications

Emotional speech feature selection using end-part segmented energy feature

Emotional speech feature selection using end-part segmented energy feature

Voice Biometrical Match of Mother and Daughter for Forensic Purpose

Automatic Speech Segmentation and Recognition using Class-Specific Features

Contact Info

Product

Resources

About