Nondestructive Determination of Maturity of the Monthong Durian by Mel-Frequency Cepstral Coefficients (MFCCs) and Neural Network

Khunarsa, Peerapol; Mahawan, Julalug; Nakjai, Pisit; Onkhum, Nerissa

doi:10.4028/www.scientific.net/amm.855.75

Cited by 3 publications

(2 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The Convolutional layer is denominated from the important linear mathematical operation in the layer, so-called convolution, or filtering. The convolution function is described in (9). ( , )…”

Section: Convolutional Neural Network For Voice Imagementioning

confidence: 99%

An Alternative Approach for Thai Automatic Speech Recognition Based on the CNN-based Keyword Spotting with Real-World Application

Sukichai¹,

Utintu²

2021

Adv. sci. technol. eng. syst. j.

View full text Add to dashboard Cite

An automatic speech recognition (ASR) is a key technology for preventing an ongoing global coronavirus epidemic. Due to the limited corpus database and the morphological diversity of the Thai language, Thai speech recognition is still difficult. In this research, the automatic speech recognition model was built differently from the traditional Thai NLP systems by using an alternative approach based on the keyword spotting (KWS) method using the Mel-frequency cepstral coefficient (MFCC) and convolutional neural network (CNN). MFCC was used in the speech feature extraction process which could convert the voice input signals into the voice feature images. Keywords on these images could then be treated as ordinary objects in the object detection domain. The YOLOv3, which is the popular CNN object detector, was proposed to localize and classify Thai keywords. The keyword spotting method was applied to categorize the Thai spontaneous spoken sentence based on the detected keywords. In order to find out the proposed technique's performance, real-world tests were carried out with three connected airport tasks. The Tiny-YOLOv3 showed the comparative results with the standard YOLOv3, thus our method could be implemented on the low-resource platform with low latency and a small memory footprint.

show abstract

Section: Convolutional Neural Network For Voice Imagementioning

confidence: 99%

An Alternative Approach for Thai Automatic Speech Recognition Based on the CNN-based Keyword Spotting with Real-World Application

Sukichai¹,

Utintu²

2021

Adv. sci. technol. eng. syst. j.

View full text Add to dashboard Cite

show abstract

“…It has been reported that audio data converted to MFCC can be effectively classified for sound classification [ 16 ], because the potentially available phonetic information is included to facilitate feature classification. Therefore, MFCCs have acoustic features that have been widely used in various applications [ 10 , 17 , 18 , 19 ]. As deep learning-based classification models have shown capabilities of high accuracy and reliability in various fields [ 20 , 21 ], the models have a high potential for effectively classifying animal vocals that contain behavioral meanings.…”

Section: Introductionmentioning

confidence: 99%

Deep Learning-Based Cattle Vocal Classification Model and Real-Time Livestock Monitoring System with Noise Filtering

Park

Kim

Moon

et al. 2021

Animals

View full text Add to dashboard Cite

The priority placed on animal welfare in the meat industry is increasing the importance of understanding livestock behavior. In this study, we developed a web-based monitoring and recording system based on artificial intelligence analysis for the classification of cattle sounds. The deep learning classification model of the system is a convolutional neural network (CNN) model that takes voice information converted to Mel-frequency cepstral coefficients (MFCCs) as input. The CNN model first achieved an accuracy of 91.38% in recognizing cattle sounds. Further, short-time Fourier transform-based noise filtering was applied to remove background noise, improving the classification model accuracy to 94.18%. Categorized cattle voices were then classified into four classes, and a total of 897 classification records were acquired for the classification model development. A final accuracy of 81.96% was obtained for the model. Our proposed web-based platform that provides information obtained from a total of 12 sound sensors provides cattle vocalization monitoring in real time, enabling farm owners to determine the status of their cattle.

show abstract

Machine Learning-Based Acoustic System for Maturity Classification of Durian Fruit Before Harvesting

Nguyen,

Huynh,

Duong

et al. 2024

Communications in Computer and Information Science

View full text Add to dashboard Cite

Nondestructive Determination of Maturity of the Monthong Durian by Mel-Frequency Cepstral Coefficients (MFCCs) and Neural Network

Cited by 3 publications

References 18 publications

An Alternative Approach for Thai Automatic Speech Recognition Based on the CNN-based Keyword Spotting with Real-World Application

An Alternative Approach for Thai Automatic Speech Recognition Based on the CNN-based Keyword Spotting with Real-World Application

Deep Learning-Based Cattle Vocal Classification Model and Real-Time Livestock Monitoring System with Noise Filtering

Machine Learning-Based Acoustic System for Maturity Classification of Durian Fruit Before Harvesting

Contact Info

Product

Resources

About