Abstract Communication barriers faced by individuals those who are deaf or voiceless pose significant challenges, particularly during emergencies when clear and rapid communication is essential. Sign language, a style of communication for many people with hearing or speech impairments, requires proficiency from both the sender and receiver. Recognizing the need for inclusive communication solutions, researchers have explored technologies that connection the breach between sign and spoken verbal communication. One approach involves vision-based technique, which utilize cameras and advanced image processing techniques to detect and understand hand movements and gestures associated with sign language. Through real-time analysis, these systems can transform sign into a spoken language called voice or text, enabling flawless communication between individuals proficient in sign verbal communication and those who are not. This technology holds immense potential for enhancing accessibility in various contexts, including emergency situations where swift and accurate communication is critical. The system obtained 98.0% accuracy, 100% recall (sensitivity), F1 score of 95%, and error analysis showed PSNR of 1% and MSE of 2%, all of which together represent excellent classification performance with some room for improvement in error robustness. The vision-based technologies aims to foster greater inclusivity and ease of access in communication for persons with disabilities, ultimately promoting a more inclusive society with a voice conversion system.
Gajjala et al. (Fri,) studied this question.