Sign Language to Voice Translator
Authors:-B. Sai Praneetha, Dyanesh R, M. Praniksha, Sanjai R J, Professor Dr. Ajay Kumar Singh
Abstract-:One area of assistive technology that is gaining popularity is its capacity to facilitate communication between people with hearing impairments and the general public. This research introduces a real-time sign language detection system that uses a single webcam to recognize the alphabet in American Sign Language (ASL) and interpret numerical gestures. Based on hand landmarks recorded by MediaPipe, the system recognizes ASL alphabets with high accuracy and recognizes digits from 1 to 10 using deep learning, computer vision, and language processing algorithms. The suggested solution combines OpenCV and MediaPipe for landmark tracking, pyttsx3 for speech feedback, and a user-friendly graphical user interface created using Tkinter. TensorFlow is used to train the alphabet identification model, while landmark distance computations and geometric logic is used to distinguish numerical movements. Because this hybrid approach guarantees real-time speed and usability, the solution is feasible for applications that are focused on accessibility, education, and assistive technology.
