Title

Sign-To-Text Conversion

AUTHOR(S)

Piyush Verma Ayush Kumar Er. Akanksha Singh

ABSTRACT

People who are born deaf or hard of hearing are viewed as liabilities by the community and find it extremely difficult to support themselves. They need specialized training in order to communicate using sign language, but the issue still exists since in India, the general public was not provided with enough sign language instruction, making it difficult for them to be understood. a method based on learning that transforms American Sign Language (ASL) into comprehensible text. Convolutional neural networks (CNNs) are used to extract identifying characteristics based on comparisons with datasets. From there, they generate predictions about the alphabet, resulting in word formations that represent the intended message. Using data processing, augmentation, and normalization approaches, the suggested model was trained on this dataset. Using a test dataset, the trained models were assessed and shown to be very accurate in identifying the ASl symbols. Through Sign-To-Text translation, this method offers a quick, accurate, and affordable way to translate American Sign Language into Text. This strategy may enhance the manner in which an individual and those with speech impairments communicate.

DOI :

Under Process

Download Full Article