Published Paper Details:

Recognizing hand gestures plays a vital role in enhancing interaction between humans and computers, particularly in assistive technologies, sign language interpretation, and real-time communication. This study explores effective methods for detecting hand gestures using Convolutional Neural Networks (CNN) and advanced deep learning techniques. The approach includes three main steps: capturing hand images using a webcam, removing noise and other distractions during image processing, and deploying a CNN architecture capable of recognizing various hand gestures, both static and dynamic. To evaluate the system's performance, a custom dataset with diverse hand gestures was created, achieving an accuracy rate exceeding 90%. Additionally, the proposed method enhances hand segmentation, accommodates gesture variability across users, and supports real-time gesture recognition, expanding the potential for interaction. The findings demonstrate that deep learning techniques significantly improve user interaction, especially for hearing-impaired individuals, by increasing recognition accuracy.