LOK: Learn Objects for Kids

Application splash screen
Application is detecting and labeling cat
Application is detecting and labeling horse
Application is detecting and labeling book
application is detecting and labeling clock
Application logo

Inspiration

The scope of application is learning and we want to enhance the learning quality of kids about the environment. This application is currently detecting objects in surrounding and give information about them we want to take it to next level and add animations using augmented reality that will develop the interest of the kids and keep their excitement at a high level. In our plans, we want to make sure if the kid is learning properly or not. We will check this by taking the voice input when a kid is pointing mobile to a specific object. The system will tell if the kid is pronouncing this object right or not. We will maintain the profile of users and add the functionality of Facebook login. Then there will be a competition between Facebook friends of the user about the learning success.

What it does

The current version of the application teaches the names of objects in English as well as Urdu. If you are using the application and it has detected some object in the surrounding, you just need to click on that object and the mobile device will say that name in English as well as Urdu. As this application is for learning purposes, we have plans to include different animations in the future that will be more exciting and interesting for kids.

How I built it

This is an Android application built using Java programming language. We used TensorFlow predefined model to detect objects in the surrounding. In the development process, we added Urdu language labels and voice for our local kids who have problems with non-native language. We have recorded voices in Urdu as well as English for limited no of objects present in the environment.

Challenges I ran into

It was a big challenge to decide the idea of making an application that is using a machine learning model for the hackathon spanned in two days and one night. But we tried and faced many difficulties developing this application. Another big challenge was we were using the TensorFlow pre-trained model for object detection. This model might be trained on data not taken from our current surroundings. It was not at its best accuracy but was working pretty well.

Accomplishments that I'm proud of

The main accomplishment we tried to develop the interest of kids in the learning process. We tried Augmented reality by labeling objects in the real world. The real-world interaction with the mobile device was a big thing that will lead us to add more functionalities to this learning process. We used a machine learning model for the learning process of kids that is in modern trends of the world.

What I learned

One of the main objective of taking part in this hackathon is learning. We have learned team skills. We have learned how to manage the work division of a project and meet deadlines. How to motivate each other for their work. One of the main things of a project development process for the hackathon is idea discussion. The better can attract more people. If a team member has an idea and it looks much better to him but the actual value of idea will be evaluated after some discussion with the team members that how this project will be beneficial for the people who will be using it. We learned that if you have a cooperative team that can compromise and value the suggestions of others then you can achieve anything. Another main part of learning in this hackathon was working with modern technologies like machine learning and adding a little feature of augmented reality. It was fun learning these new things. We didn't that much idea about these things but as the best way of learning, something is by doing it.