Gesture Voice: Transforming Gestures into Text Transcription

With the explosive growth of video content and the pressing necessity of accessible communication, particularly for the deaf and hard-of-hearing, automated systems that can interpret sign language from video to coherent, contextually relevant sentences have become a necessity. This paper introduces a strong, end-to-end system that makes use of YOLO-based gesture detection, Roboflow-based custom dataset annotation, and advanced NLP modules to interpret sign language video input into structured, human-understandable text. In contrast to earlier techniques bound by hardware constraints, static gesture recognition, or contextual insensitivity, our system exhibits higher accuracy, real-time processing, and flexibility. Through rigorous comparative analysis with the best state-of-the-art methods, we establish noteworthy improvements in recognition accuracy, sentence generation, and usability. Exhaustive experiments establish the efficacy of the system across varied scenarios, establishing a new benchmark in accessibility, education, and automated video summarization.