Continuous ASL Translation
In a world where communication is integral to opportunities and relationships, the hearing-impaired face significant challenges due to communication barriers. This project tackles the lack of popularity of sign language, creating a solution that automatically translates Sign Language gestures into spoken language. Leveraging Artificial Intelligence (AI) fields like Machine Learning, Deep Learning, Computer Vision, Object Detection, Object Recognition, and Natural Language Processing, the system aims to empower mute and deaf individuals. The initiative aligns with a wealth of related studies emphasizing the importance of sign language recognition and translation. Addressing the global prevalence of hearing loss, the project employs innovative approaches, including the Transformer and Conformer architectures. The models aim to enhance robustness and versatility through extensive preprocessing and data augmentation. The article structure covers background, methodology, challenges, acquired results, and unexplored techniques. This endeavour strives to break down communication barriers, fostering inclusivity for the hearing-impaired in our communication-dependent world.
- Python,
- Conformer,
- Transformers,
- Deep Learning
