Sign Talk Assistive Technology: Real-Time Recognition of the Libras Typical Alphabet Using Artificial Intelligence

Objective: The objective of this research is to promote SignTalk, an innovation in assistive technology based on artificial intelligence that recognizes the finger alphabet of Brazilian Sign Language in real time through a webcam and programming. Theoretical Framework: With this innovation, the future of communication through Libras becomes easier, more accessible and faster. Inclusive communication is being shaped by a convergence of disciplines that seek to facilitate interaction and communication between deaf and non-deaf people. Method: The method used included publications on assistive technologies and similar technologies that made it possible to develop SignTalk using the Python language and machine learning with thirteen thousand images. Results and Discussion: The results indicate that in the beta version, SignTalk has shown great inclusive capacity, being a teaching and learning tool of the finger alphabet for the deaf public. Implications of the Research: This innovation is an inclusive communication tool that uses the guidelines of Sign Language to contribute to the inclusive education of deaf people. Originality/Value: The study of gestures and machine learning innovates by enabling the system to recognize human hand gestures, returning the corresponding graphical representation of fingers quickly and accurately.