Inspiration

Making sign language accessible for all! 1 in 5 people (or 12 million people) in the UK are Deaf or hard of hearing Only 1% of the hearing-impaired and deaf population in the United States had the resources to learn and use sign language The UK are in the process of offering a sign language GCSE Our tool would be the perfect compliment to the lessons to set the students homework

What it does

How do students practice and test their sign language efficiently when most likely no one else in their household knows it? How can teachers monitor whether students are doing their homework and how they are performing on it? ASLearn solves these problems Can be used all across the country, teacher sets the class and so the level of difficulty and content can be edited by them to suit any level of student Image classification based on the children’s actions

How we built it

Delegated tasks based on specialisations and areas of interest Found ASL datasets, trained models but the datasets did not have enough variety so they did not work on images we took. So, we made our own dataset and trained it on the model. Meanwhile Daniel built the teacher and student sides of the website. Then we linked it all together to get our final product.

Challenges we ran into

Datasets for training an effective neural network so we created our own.

Accomplishments that we're proud of

Educating more people on sign language opens up the world for the hearing impaired Gives more deaf people an effective communication skill Gives school children additional skills which may help improve their employability Allows teachers to effectively set work for the students based on data on how they performed which will optimise their learning

What we learned

Importance of variety in a dataset. Sign Language Recognition is much more complicated than we first anticipated.

What's next for ASLearn

More lessons, challenges! Set level to suit students.

Built With

Share this project:

Updates