Classification of Sign Language in Real Time Using Convolutional Neural Network

Abstract

Communication between people is essential for daily life activities. However, humans are created with their own strengths and weaknesses. One of them is the difficulty of communication and interaction for people with hearing and speech impairments. Sign language is a language for people who have difficulty hearing and speaking. However, sign language is not popular in society, and people who have it will have more difficulties. This research aims to classify hand gestures of sign language into letters using a convolutional neural network (CNN). The dataset is obtained from Kaggle, with a total of 34,627 data divided by the ratio of training and testing data of 80:20. From the test results, the letters of the alphabet that can be translated are: A, B, C, D, E, F, G, H, I, J, K, L, M, N, O, P, Q, S, T, U, V, W, X, Y, and Z. Furthermore, validation accuracy is obtained. In this study, a very high validation accuracy was obtained. The easiest letters to guess are V and N, while the most difficult letters to guess are n, c, j, and z. With different preprocessing, the loss value can be reduced, giving a higher accuracy of 95.4%.

Mask Detection App Uses Haar Cascade and Convolutional Neural Network to Alert Comply with Health Protocols

“…Each feature map is followed by a nonlinear function, such as the Rectified Linear Unit (ReLU) nonlinearity, which is popular because of its ease of calculation. The pooling layer minimizes the sample count [13].…”

K-Means Clustering and Multilayer Perceptron for Categorizing Student Business Groups

“…Subsequently, compared to the polynomial regression method, MLP shows better performance [27]. Python was chosen as the programming language for this study since it is considered one of the easiest to learn and utilize [28]. A machine learning model developed from this study can execute data target labeling and categorization with the highest precision and accuracy.…”