Convolution Neural Network for Predicting Alphabet Sign Language and Comparative Performance Analysis of CNN, KNN, and SVM Algorithms

The use of sign language is a crucial component of daily life for those who have trouble speaking or hearing. American Sign Language is challenging for computer vision to recognise due to its increasing complexity and high intraclass diversity. This research compares and evaluates the effectiveness of three machine learning algorithms in the task of sign language detection. By utilising the Convolution Neural Network, K-Nearest Neighbor, and Support Vector Machine algorithms, we hope to represent characteristics that will be learnt. After that, we must evaluate them in terms of their accuracy, recall, precision, F-Value, and mean absolute percentage error (MAPE). We need to determine which algorithm is most suited for classifying sign language images after taking into account all of these factors. Additionally, we are employing the CNN algorithm to recognise sign language in both text and audio.