Enhancements in Continuous Kannada ASR System by Background Noise Elimination

G, Thimmaraja Yadava; Nagaraja, B G; Jayanna, H. S.

doi:10.1007/s00034-022-01973-0

Cited by 7 publications

(5 citation statements)

References 26 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Using continuous Kannada speech data transcription, phoneme set, lexicon, silence phones and Kaldi recipe, the achieved WER are shown in Table 1. From the Table, it can be observed that the hybrid modeling technique, DNN+HMM is outperformed the other acoustic modelling techniques in terms of WER and has given a better accuracy of speech recognition compared to the existing works (15,16,20) . This is since the vocabulary used for the system training and testing is less and the outcome of the triphone 3 is given as input to the DNN.…”

Section: Resultsmentioning

confidence: 97%

“…The decoding results show the recognition of continuous Kannada speech sentences which is represented inFigure 6. (15) WER achieved in (16) WER achieved in (20) Achieved WER in % for small vocabulary:…”

Section: Resultsmentioning

confidence: 99%

“…The authors have achieved an accuracy of 95.9% for noisy speech data and they have used this model in SQS. The authors further enhanced the recognition accuracy of continuous Kannada SQS by integrating the noise reduction algorithm before speech feature extraction (16) . An ensemble of proposed noise elimination algorithm (17) and time delay neural network (TDNN) acoustic modelling technique has given a better improvement and achieved an accuracy of 97.60% compared to earlier continuous Kannada SQS (15) .…”

Section: Introductionmentioning

confidence: 99%

“…The authors have achieved a WER of 4.64% using DNN-HMM. All our previous works (11)(12)(13)(14)(15)(16)(17) described the development of large vocabulary isolated and continuous E2E ASR systems for Kannada language/dialects. In this work, an attempt is made to develop a continuous speech-to-text system for small vocabulary for Kannada language/dialects.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Development of Small Vocabulary Continuous Speech-to-Text System for Kannada Language/Dialects

G¹,

Nagaraja²,

Kumaran³

et al. 2022

IJST

View full text Add to dashboard Cite

Objectives: To develop a speech-to-text (STT) system using Kaldi speech recognition toolkit for continuous Kannada language/dialects. Methods: A continuous Kannada speech data is collected from 100 speakers/farmers of Karnataka state in field. The lexicon/dictionary and set of phonemes for Kannada language/dialects are created and transcribed the collected speech data using transcriber tool. The ASR models are developed at different phoneme levels using Kaldi. Findings: In this work, an effort is made to develop a robust small vocabulary STT system for continuous Kannada language using Kaldi. The various acoustic modelling techniques are used to develop a robust ASR model and achieved a word error rate (WER) of 0.23%. The performance of the developed ASR model is compared with existing works and analyzed by offline speech recognition. Novelty: Many STT systems have been developed for Indian and International languages/dialects, but not for Kannada language. This work is first of its kind using Kaldi in Kannada language under the constraints of limited data. The developed ASR model could be used further in the development of end-to-end ASR system for speech processing applications.

show abstract

Section: Resultsmentioning

confidence: 97%

Section: Resultsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Development of Small Vocabulary Continuous Speech-to-Text System for Kannada Language/Dialects

G¹,

Nagaraja²,

Kumaran³

et al. 2022

IJST

View full text Add to dashboard Cite

show abstract

“…The works in (6,7) focus on recognizing spoken isolated words in Kannada. ASR for continuous Kannada speech has been developed by the studies in (8)(9)(10) . Achieving an accuracy of 90% or more for a continuous ASR requires sufficient amount of training data.…”

Section: Introductionmentioning

confidence: 99%

Low Resource Kannada Speech Recognition using Lattice Rescoring and Speech Synthesis

Murthy¹,

Sitaram²

2023

IJST

View full text Add to dashboard Cite

Objectives: Improving the accuracy of low resource speech recognition in a model trained on only 4 hours of transcribed continuous speech in Kannada language, using data augmentation. Methods: Baseline language model is augmented with unigram counts of words, that are present in the Wikipedia text corpus but absent in the baseline, for initial decoding. Lattice rescoring is then applied using the language model augmented with Wikipedia text. Speech synthesis-based augmentation with multi-speaker syllable-based synthesis, using voices in Kannada and cross-lingual Telugu languages, is employed. We synthesize basic syllables, syllables with consonant conjuncts, and words that contain syllables that are absent in the training speech, for Kannada language. Findings: An overall word error rate (WER) of 9.04% is achieved over a baseline WER of 40.93%. Language model augmentation and lattice rescoring gives an absolute improvement of 16.68%. Applying our method of syllable-based speech synthesis over language model augmentation and rescoring yields a total reduction of 31.89% in WER. The proposed approach of language model augmentation is memory efficient and consumes only 1/8 th the memory required for decoding with Wikipedia augmented language model (2 gigabytes versus 18 gigabytes) while giving comparable WER (22.95% for Wikipedia versus 24.25% for our method). Augmentation with synthesized syllables enhances the ability of the speech recognition model to recognize basic sounds thus improving recognition of out-of-vocabulary words to 90% and in-vocabulary words to 97%. Novelty: We propose novel methods of language model augmentation and synthesis-based augmentation to achieve low WER for a speech recognition model trained on only 4 hours of continuous speech. Obtaining high recognition accuracy (or low WER) for very small speech corpus is a challenge. In this paper, we demonstrate that high accuracy can be achieved using data augmentation for a small corpus-based speech recognition.

show abstract

Advancements in encoded speech data by background noise suppression under uncontrolled environment

Nagaraja,

Yadava,

Anees

2024

Int J Speech Technol

View full text Add to dashboard Cite

Enhancements in Continuous Kannada ASR System by Background Noise Elimination

Cited by 7 publications

References 26 publications

Development of Small Vocabulary Continuous Speech-to-Text System for Kannada Language/Dialects

Development of Small Vocabulary Continuous Speech-to-Text System for Kannada Language/Dialects

Low Resource Kannada Speech Recognition using Lattice Rescoring and Speech Synthesis

Advancements in encoded speech data by background noise suppression under uncontrolled environment

Contact Info

Product

Resources

About