International Journal of Scientific Research and Engineering Development

Communication Between Individuals Who Use American Sign Language (Asl) And People Who Do Not Understand Sign Language Remains A Significant Accessibility Challenge. This Paper Presents Signfusion, A Real-Time, Bidirectional Asl Communication System Designed To Bridge This Communication Gap Using Computer Vision, Machine Learning, Speech Processing, And Web Technologies. The System Is Implemented Using Python, Opencv, Mediapipe, Tensorflow/Keras, ScikitLearn, Speechrecognition, Pyttsx3/Gtts, Flask, And Socket.Io. Experimental Evaluation Reported 97.3% Recognition Accuracy On The Standard Asl Alphabet Test Dataset And Approximately 91.6% Recognition Accuracy Under Live Conditions. The Random Forest Pipeline Achieved An Average FrameTo-Dashboard Latency Of 56 Ms, While The System Maintained Real-Time Operation At Approximately 28–32 Frames Per Second On Standard Hardware. These Results Demonstrate The Feasibility Of A Software-Only, Low-Cost System For Real-Time Bidirectional Asl Communication.

Sushma C K, "A Real-Time Bidirectional American Sign Language Communication System Using Mediapipe and Deep Learning" International Journal of Scientific Research and Engineering Development, V9(4): Page(1191-1195) July-August 2026. ISSN: 2581-7175. www.ijsred.com. Published by Scientific and Academic Research Publishing.