Treffer: Development of a Speech-to-Sign Language Translation System Using Machine Learning and Computer Vision: A Bulgarian Case Study.

While sign language is the primary communication tool for deaf individuals, the limited number of sign language interpreters creates substantial obstacles to social integration and access to essential services. The paper presents the development of an automated translation system that converts spoken Bulgarian language into animated Bulgarian Sign Language (BgSL) using machine learning and computer vision techniques. The system addresses the communication barrier between hearing and deaf individuals by providing real-time translation capabilities. The proposed solution combines speech recognition, natural language processing, and videobased gesture visualization to create an accessible communication tool. The system achieves 90% accuracy in speech recognition and 78% accuracy in text-to-gesture correspondence, demonstrating its potential for practical applications. The paper discusses future improvements and potential applications in education. [ABSTRACT FROM AUTHOR]