Real-Time American Sign Language Detection Using YOLOv11: An Enhanced Deep Learning Approach

Recent advances in deep learning have revolutionized the field of sign language recognition, yet real-time detection remains challenging. This chapter introduces a novel approach to American Sign Language (ASL) detection using YOLOv11, incorporating Advanced Programmable Gradient Information (APGI) and Enhanced Generalized Efficient Layer Aggregation Network (E-GELAN). Our comprehensive system achieves 98.2% detection accuracy while maintaining real-time performance at 60 frames per second on consumer-grade Graphics Processing Unit. Through extensive experimentation, we demonstrate a 15% improvement in accuracy and 30% reduction in inference time compared to previous approaches reported, making our solution particularly suitable for real-world applications in assistive technology. The proposed architecture effectively addresses the challenges of varying lighting conditions, user distances, and sign variations, representing a significant step forward in accessible communication technology.