About this document
Two-Hand Arabic Sign Language Recognition by ghous97788 is a document available to read on EtoBox.
The document presents a two-hand sign language recognition system for Arabic Sign Language (ArSL) that addresses challenges such as signer variability and occlusions by integrating keypoint-based descriptors and shape features with an attention-driven feature fusion strategy. The system demonstrates robust performance with accuracy rates of 77.34%, 62.53%, and 47.58% on various datasets, showcasing its effectiveness in recognizing both static and dynamic gestures. The proposed methodology emphasizes the imp
- Author
- ghous97788
- Language
- EN