SIGN LANGUAGE TRANSLATOR FOR SPEECH-IMPAIRED
M. Muktheswar Reddy1, N. Madhu2, M. Bhagavathi3, K. Tanvi4, Mr. U. Nagaiah5
1,2,3,4Department of CSE (AI & ML), CMR Engineering College, Hyderabad.
5Assistant Professor, Department of CSE (AI & ML), CMR Engineering College, Hyderabad.
Abstract: This project aims to develop an innovative Sign Language Translator for individuals with speech impairments, utilizing the power of OpenCV and Convolutional Neural Networks (CNNs). The communication challenges faced by speech-impaired individuals often lead to barriers in their interactions with the hearing world, making an effective sign language translation system essential. To address this need, we employ OpenCV for real-time video processing, allowing the system to capture and analyze sign language gestures from a live video feed. The core of this technology is the CNN-based model, meticulously trained on a comprehensive dataset of sign language gestures, enabling it to accurately recognize and translate these gestures into spoken language. The system's workflow involves capturing video frames from the input source, passing them through the trained CNN model, and subsequently converting the recognized signs into textual representations. These translated signs can be combined to form complete sentences or phrases in spoken language, providing a seamless and intuitive means of communication for speech-impaired individuals. Additionally, the system can display the recognized signs as text on the screen, facilitating visual confirmation for users. Moreover, this technology's adaptability allows for training the CNN model on specific sign language datasets, making it customizable to different sign languages. By harnessing the capabilities of OpenCV and CNNs, the Sign Language Translator offers an opportunity to enhance communication, break down communication barriers, and promote inclusivity for the speech-impaired community, ultimately fostering more effective interactions with the hearing world.
Keyword: American Sign Language Convolutional Neural Network, Deep learning, gesture recognition, hand gesture to speech.