Upload Image
Upload and Test
Results
AR Models
Exploration of AR Models

Enciclopedia de Animales (Animal Encyclopedia)

💡 Inspiration

Encyclopedias are highly recommended as a starting point for your research on a particular topic. Encyclopedias will give you introductory information to help you broaden or narrow your topic, while also providing keywords and terms needed to conduct further research. An AR-based encyclopedia can help them to understand things faster and in a more efficient way.

💻 What it does

Enciclopedia de Animales is an AR-based Animal Encyclopedia, that elucidates their features, and imitates them into the real world using AR (Augmented Reality) technology. It also features a CNN that could identify animals from their images.

⚙️ How we built it

ML: Python, TensorFlow
Frontend: HTML, CSS, JS, Three Js
Backend: Django
Database: CockroachDB
Authentication: Auth0
Sending fun facts: Twilio
AR: echoAR

About the ML Model

Dataset had images of 10 different animals
Around 2.5K images per animal
There were 5 Convolutional Layers in our model and 2 Dense layers followed by an output layer
The final accuracy is around 82%
The saved model is compact and around just 1.5 MB in size (ready-to-use in low-end compact devices)

Animals for which the ML model works -

dog
horse
elephant
butterfly
chicken
cat
cow
sheep
spider
squirrel

Use of Twilio

For sending the facts to the email used for registration.

Use of CockroachDB

We have used CockroachDB as a primary database because it is an easy-to-use, open-source, and indestructible SQL database.

🔑 Auth0

We have used Auth0 for secure user authentication

🧠 Challenges we ran into

One challenging part was how to implement the AR and rendering the 3D models using react-three-fiber, and it took a while to learn and figure out how to make it work properly on the website.
Working on an ML model that could classify so many animals was no party either and took a major chunk of our time.