Kumoro, Calvindra Laksmono (2024) Pengembangan Aplikasi Berbasis Website Untuk Klasifikasi Tangisan Bayi Dengan Pendekatan Deep Learning. Other thesis, Institut Teknologi Sepuluh Nopember.
![[thumbnail of 5027201020-Undergraduate_Thesis.pdf]](http://repository.its.ac.id/style/images/fileicons/text.png) |
Text
5027201020-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 July 2026. Download (1MB) | Request a copy |
Abstract
Komunikasi adalah cara penyampaian ide, pandangan, dan pesan melalui berbagai medium seperti kata-kata, gerakan, gambar, dan simbol-simbol. Komunikasi dilakukan oleh semua makhluk hidup dengan caranya masing-masing seperti yang dilakukan pada bayi dengan cara menangis. Menangis merupakan penyampaian kebutuhan dan perasaan bayi kepada orang tua namun tangisan tersebut seringkali sulit untuk dipahami, karena terdengar serupa. Ketidakpahaman terhadap tangisan bayi ini dapat menyebabkan masalah seperti bertambah kencang dan lamanya tangisan bayi yang mampu mempengaruhi kesehatan fisik dan mental orang tua. Oleh karena itu, penelitian ini mengusulkan pengembangan web apps yang dapat membantu mengklasifikasi tangisan bayi dengan lebih baik. Pengembangan aplikasi ini menggunakan metode deep learning dengan memanfaatkan algoritma Mel Frequency Cepstral Coefficients (MFCC) untuk ekstraksi fitur yang digunakan dalam pengolahan sinyal suara, dan Convolutional Neural Network (CNN) untuk mengklasifikasi data audio. Diantara dataset pertama, dataset kedua yang sudah diaugmentasi, dan dataset ketiga (dataset kedua yang di-undersampling), didapatkan hasil f1-score tertinggi pada dataset kedua pada masing-masing kelas. Pada dataset kedua masing-masing kelas didapat f1-score belly pain sebesar 64%, burping 59%, discomfort 57%, hungry 64%, tired 50%. Implementasi web apps pada penelitian ini mampu melakukan klasifikasi suara dengan fitur upload file dan record audio. Audio dengan kondisi kebisingan yang berbeda, dan terdapat bagian audio yang terpotong suara tangisannya dapat mempengaruhi tingkat akurasi dari model CNN
=====================================================================================================================================
Communication is a way of conveying ideas, views, and messages through various mediums such as words, gestures, pictures, and symbols. Communication is carried out by all living things in their own way, as babies do by crying. Crying is the delivery of the baby's needs and feelings to the parents but the cries are often difficult to understand, because they sound similar. Not understanding the baby's cry can cause problems such as increasing the loudness and length of the baby's cry which can affect the physical and mental health of the parents. Therefore, this research proposes the development of web apps that can help classify baby cries better. This application development uses deep learning method by utilizing Mel Frequency Cepstral Coefficients (MFCC) algorithm for feature extraction used in sound signal processing, and Convolutional Neural Network (CNN) to classify audio data. Among the first dataset, the segmented second dataset, and the third dataset (the second dataset that was under sampled), the highest f1-score results were obtained in the second dataset in each class. In the second dataset for each class, the f1-score of belly pain is 64%, burping 59%, discomfort 57%, hungry 64%, tired 50%. The implementation of web apps in this study can perform voice classification with file upload and audio record features. Audio with different noise conditions, and there are parts of the audio that are cut off by crying sounds can affect the accuracy of the CNN model
| Item Type: | Thesis (Other) |
|---|---|
| Uncontrolled Keywords: | Klasifikasi Tangisan Bayi, MFCC, CNN, Website, Mobile Apps; Baby Crying Classification, MFCC, CNN, Website, Mobile Apps. |
| Subjects: | T Technology > T Technology (General) > T385 Visualization--Technique T Technology > T Technology (General) > T57.5 Data Processing T Technology > T Technology (General) > T58.8 Productivity. Efficiency |
| Divisions: | Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Information Technology > 59201-(S1) Undergraduate Thesis |
| Depositing User: | Calvindra Laksmono Kumoro |
| Date Deposited: | 01 Feb 2024 06:01 |
| Last Modified: | 01 Feb 2024 06:01 |
| URI: | http://repository.its.ac.id/id/eprint/105891 |
Actions (login required)
 |
View Item |