Communication barriers between deaf, mute individuals and the hearing majority hinder participation in education and daily life. Existing assistive technologies often provide one-way translation, limiting reciprocal interaction and reducing inclusivity. This study developed “Signslator”, a two-way communication system integrating artificial intelligence (AI) and image processing. The prototype utilized a Raspberry Pi 5, camera, microphone, dual monitors, and Python-based software with MediaPipe for gesture recognition. A dataset of 150 school-related words, validated by Special Education (SPED) teachers, was used for training. System performance was tested under controlled conditions with varying confidence thresholds, and evaluated using accuracy, precision, recall, F1-score, and weighted mean ratings from expert feedback. The system achieved 100% accuracy, precision, recall, and F1-score in recognizing letters (A–Z) and numbers (1–9). For the 150-word vocabulary, average recognition accuracy was 84.93%, with 130 words scoring above 50%. However, 20 words fell below 49% accuracy, highlighting limitations in complex or visually similar gestures. Expert evaluation rated the system highly functional (AWM = 4.25), reliable (AWM = 4.21), usable (AWM = 4.55), and efficient (AWM = 4.33). The AI-driven auto-correction feature improved clarity by merging fragmented inputs into coherent words. Findings confirm that “Signslator” effectively bridges communication gaps in classroom settings, enabling real-time two-way translation between spoken language and American Sign Language. While highly accurate for basic gestures, future improvements should expand vocabulary, incorporate facial expressions, and optimize performance under diverse environmental conditions. The system demonstrates strong potential as an inclusive educational tool.

Keywords: Deaf Individuals, Sign Language Translation, Hand Gesture Recognition, Artificial Intelligence in Education, Image Processing, Assistive Technology, Two-Way Communication