Apines, Ronald Jr., S. <br/><br/>
Sign gesture recognition system / 
Ronald S. Apines, Jr., Jonas Andrei C. Ballarta, Aerom Von J. Canimo, Ray John  B. Cantonjos. 
 - 100 pages :  illustrations ;  28 cm. +  1 CD-ROM (4 3/4 in.) 
<br/><br/> Thesis (Undergraduate) 
<br/><br/>College of Science -- 
<br/><br/>Includes bibliographical references. 
<br/><br/> Speech is the most prevalent means of communication. This communication method, however, is impractical for the deaf community, who interact through sign language. For many years, there has undoubtedly been a communication barrier between signers and non-signers. The necessities of communities and rapidly evolving technologies inspire researchers to devise innovative solutions to meet these demands. Researchers conduct studies, particularly in the field of human-computer interaction (HCI). The Sign Language Recognition System (SLR) is widely valued because it bridges the gap between signers and non-signers. This study applied a deep learning computer vision method to recognize hand gestures by establishing an Artificial Neural Network architecture model (Long short-term memory). The model learns to identify hand gesture photos over an epoch. The Sign Gesture Recognition System was developed using TensorFlow lite, an open-source framework for deep learning with on-device inference. OpenCV for real-time computer vision was utilized to establish the proposed Application successfully. Python, Media Pipe, LSTM, TensorFlow, and Keras for hand gesture recognition., VS Code for an Integrated Development Environment (IDE), a scripting language for application development, and open-source web technologies including HTML, JavaScript, and CSS to design and build the Application. It all starts with data acquisition. Following that is hand gesture recognition; once the model recognizes the gesture, a corresponding word or letter will be displayed on the screen. Using a confusion matrix and ISO 25010, the researchers evaluated all procedures to produce accurate and trustworthy results. The system was assessed as "Moderately Reliable" by ten signers. --Author's Abstract. 
<br/><br/><label>Subjects--Topical Terms: </label><br/>Gesture--Data processing.<br/>American sign language.<br/>Human-computer interaction.
<br/><br/><label>Subjects--Index Terms: </label>Hand gesture recognition. Sign language recognition. 
<br/><br/><label>LC Class. No.: </label>BTH QA 76  / A65 2022