Speech Recognition Using Convolutional Neural Networks

Nagajyothi, D.; Siddaiah, P.

doi:10.14419/ijet.v7i4.6.20449

Cited by 33 publications

(9 citation statements)

References 5 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Meskipun CNN dirancang untuk digunakan pada pengenalan objek dalam gambar, tetapi jika dibandingkan dengan algoritma klasik machine learning lainnya, CNN memiliki kemampuan yang lebih baik dalam mendeteksi emosi [10]. CNN dapat bekerja dengan baik apabila diberi banyak sinyal seperti perubahan frekuensi dan perubahan amplitudo yang ditemukan di dalam speech [11].…”

Section: Tinjauan Pustaka Convolutional Neural Network (Cnn)unclassified

Deteksi Emosi Berdasarkan Wicara Menggunakan Deep Learning Model

Rahmadani

Rahayu

Salim

et al. 2022

JINTEKS

View full text Add to dashboard Cite

Kemampuan komputer dalam meniru kemapuan manusia sampai saat ini menjadi hal yang menarik untuk dikembangkan. Di beberapa studi, pengenalan emosi sudah banyak diteliti baik melalui foto wajah dan ucapan verbal maupun non verbal. Penelitian ini bertujuan untuk mengeksplorasi berbagai metode Deep Learning untuk mendapatkan model yang paling baik dalam mendeteksi emosi menggunakan dataset EmoDB. Feature extraction dilakukan dengan menggunakan Zero Crossing Rate, Chroma_stft, Mel Frequency Cepstral Coefficients (MFCC), Root Mean Square (RMS) dan MelSpectogram. Pada tahap preprocessing diterapkan teknik data augmentation dengan mengaplikasikan noise injection, shifting time dan mengubah pitch dan kecepatan audio. Dari hasil penelitian dikemukakan bahwa metode deep learning terbaik berdasarkan nilai akurasi adalah CNN-BiLSTM.

show abstract

Section: Tinjauan Pustaka Convolutional Neural Network (Cnn)unclassified

Deteksi Emosi Berdasarkan Wicara Menggunakan Deep Learning Model

Rahmadani

Rahayu

Salim

et al. 2022

JINTEKS

View full text Add to dashboard Cite

show abstract

“…The CNN exploits domain knowledge about feature invariances within its structure [6]. Recently CNN has met a significant research progress [7][8][9]. A convolutional neural network consists of an input layer, hidden layers, and an output layer.…”

Section: Speech Recognition With Cnnmentioning

confidence: 99%

Early Stages of Automatic Speech Recognition (ASR) in Non-english Speaking Countries and Factors That Affect the Recognition Process

Yousufzai¹

2021

AJNNA

View full text Add to dashboard Cite

There has been a considerable stream in ASR over the past few decades, but it may seem strange why this field is still a subject for researchers to work on. There are many reasons, but somewhat because the discipline is created with the promise of human-level performance under pragmatic states and this is an inextricable problem. In addition, the increasing advancement of technology in various fields has caused a more compelling need for this field. Especially the establishment of such a system in the security sector in insecure third world countries such as Afghanistan is an urgent need. This paper began with the reflection of all the necessary knowledge about speech recognition and then suggested an unprecedented method for building an automated speech recognition (ASR) system in the Dari language using the two most powerful open source engines CMUSphinx, from Carnegie Mellon University and DeepSpeech v0.9.3 /. These systems are much more impressive than early speech recognition systems. Using my own collected dataset, a speech-to-text model has been trained for the Dari language. Firstly, the dataset is filtered according to the task, then demonstrated the possible compatibility from the hidden Markov (HMM) models, the phoneme concept to RNN training. The system surpassed previously predicted results, as CMUSphinx stated, "for a typical 10-hour operation, the WER should be around 10%." Finally, 3.3% WER was achieved with 10.3-hours of audio recording using CMUSphinx. 1% WER with DeepSpeech.

show abstract

“…If the word was spoken correctly, the game will move to the next word else further analysis of the speech input is made to check where the child is going wrong. Automatic Speech Recognition is a method that converts speech to text using the features extracted in the previous stage and Neural Networks [9].…”

Section: Automatic Speech Recognition (Asr)mentioning

confidence: 99%

A Game Application to assist Speech Language Pathologists in the Assessment of Children with Speech Disorders

2020

IJATCSE

View full text Add to dashboard Cite

Speech Disorders in children is the fifth highest occurring disability in India. Speech Disorders have a prevalence of 8 to 9 percent in children. Thus early diagnosis and treatment can increase the recovery rate among children in the age group of 3-12 years. Due to the shortage in the number of Speech Language Pathologists all over world, it becomes impossible for everyone to seek treatment. This brought forth a need for alternative modes of Speech Therapy. Computer-Based Speech Therapy is the most preferred alternative mode of speech therapy due to its availability and effectiveness. But most Computer Based applications are very expensive and cannot be afforded by the common man. In this paper we introduce a game application which assesses the speech of a child by analyzing the voice activity and the mispronounced phonemes in the speech. This application is aimed to provide assistance to Speech Language Pathologists.

show abstract

Speech Recognition Using Convolutional Neural Networks

Cited by 33 publications

References 5 publications

Deteksi Emosi Berdasarkan Wicara Menggunakan Deep Learning Model

Deteksi Emosi Berdasarkan Wicara Menggunakan Deep Learning Model

Early Stages of Automatic Speech Recognition (ASR) in Non-english Speaking Countries and Factors That Affect the Recognition Process

A Game Application to assist Speech Language Pathologists in the Assessment of Children with Speech Disorders

Contact Info

Product

Resources

About