About Ishaara
Our mission is to break down communication barriers and build a more inclusive digital world using cutting-edge, on-device Artificial Intelligence.
Ishaara is a production-grade AI system designed to translate Indian Sign Language (ISL) into text in real time using computer vision and machine learning. Built with scalability and accessibility in mind, it runs directly in the browser without requiring specialized hardware.
Over 63 million people in India, or about 6.3% of the population, are hard of hearing. This group is expected to grow due to factors such as aging. Though they do not regard their hearing loss as a disability, but rather a different way of life, their social interactions can be significantly limited.
Our project focuses on creating a robust, scalable system to predict, interpret, and translate Indian Sign Language (ISL) in real-time, without requiring specialized hardware. ISL combines actions, facial expressions, and body language, differing from other sign languages like American Sign Language (ASL), which may use single-hand gestures; ISL usually uses both hands. This complexity presents challenges in developing an accurate machine learning model for ISL interpretation.
We annotated images by capturing them from various angles, about 23k in total and then augmented the data to create a more generalized model. This solution operates entirely without a backend, making it lightweight and highly efficient.
Who Built Ishaara?
Ishaara was developed by a dedicated team of engineers: Rehan Sayyed, Vivek Chouhan, Nishikant Raut, and Rohit Deshmukh. We are united by a shared goal of making communication accessible through AI-powered Indian Sign Language translation.
Frequently Asked Questions
Q: Who created Ishaara?
A: Ishaara was developed by a team of engineers: Rehan Sayyed, Vivek Chouhan, Nishikant Raut, and Rohit Deshmukh.
Q: What is Ishaara?
A: Ishaara is an AI-powered Indian Sign Language translator that converts hand gestures into text in real time.