Sign Language Recognition
A design is presented that can recognize various American sign language static hand motions in real-time using transfer learning, Python, and OpenCV and recognize “Hello, Yes, No, Thank You, and I Love You" are all prevalent sign language terms that the system correctly acknowledges.
Abstract
Abstract: People communicate using sign language by visually conveying sign patterns to portray purpose. One method of communicating with deaf-mute people is to use sign language mechanisms. One of the nonverbal communication strategies used in sign language is the hand gesture. Many manufacturers all over the world have created various sign language systems, but they are neither adaptable nor cost-effective for end users. We present a design that can recognize various American sign language static hand motions in real-time using transfer learning, Python, and OpenCV in this paper. “Hello, Yes, No, Thank You, and I Love You" are all prevalent sign language terms that our system correctly acknowledges. The following are the key steps in system design; we created our own dataset taking prominent gestures of the American Sign Language, captured images with OpenCV and webcam, the images were then labelled for object detection, training and testing of dataset was done with transfer learning using SSD MobileNet, and eventually the gestures were successfully determined in real-time.