Little Signs
A new and Interactive way to learn Indian Sign Language
Created on 22nd February 2025
•
Little Signs
A new and Interactive way to learn Indian Sign Language
The problem Little Signs solves
According to a WHO estimate, there are about 63 million people in India suffering from significant auditory impairment. They can't speak and listen like a normal human being do. So they use special symbols to communicate, for Indian language the collection of these symbols is called as Inidan Sign Language. Till now there exist no official application for learning these symbols. There is a website where we can go through these but that is unorganised and not managed properly also the loading time is too high for such a website that a normal student who wants to learn ISL can't refer to that or he/she must have to spend a lot of time for finding a proper way to get data and learn from that website. My main aim is to resolve this issue by providing an application. Introducing LITTLE SIGNS in which as of now there are around 250+ commonly used words that are categorised into 13 Categories that provides student a interactive way of learning ISL. The application is completly offline and easy to use. The gamification of all these makes the application more interactive for the students especially. The future scope would be like increasing the categories and number of words along with sentences and adding personalised learnings through AI.
Challenges I ran into
Another challenge was to find a dataset for my application that I found on data.gov.in where I used a free to use and modify dataset provided by Government of India. The dataset name is: "Indian Sign Language Dictionary till January 2024."
Another main challange was to integrate a huge number of videos into a single application with consideration of the app size. As of now the application has 250+ videos that size about 15 MB by applying commpression at 3 different levels. For Example: Videos I am using are from a Open Government Data Repository and round of 2 MB each. Firstly I compressed using the simple editor software by decreasing the resolution to 540p that comes to around 1 MB. Then I used Handbrake an Industry standard compression software It gave me size around 250KB then I used an online compression tool that gave me size of around 30-35KB this allowed me to compress of 500MB to 15 MB.
Another challenge was to find a way to reduced the number of learning screens as there were 13 different categories in the application so I've to create 13 separate screens but I resolved it by using the Object Oriented approach and result of that I've to develop just 3 screens for these 13 categories.
Technologies used
