Wildlife Watch

Inspiration

The importance of protecting wildlife has never been more critical. Habitat loss, poaching, and climate change threatening biodiversity. It is necessary to raise the awareness about endangered species and improve the public engagement with nature. A small step to apply AI to make an impact for the nature!!!

What it does

The web app allows users to take photos of wildlife and upload it. The app will then utilise Google Vision and ChatGPT-4o to identify and provide information about the animal, including their conservation status, diet, diurnality etc. It enables users to identify endangered animals, take appropriate action if it's dangerous, and identify abnormal behaviour patterns.

How we built it

The frontend is built with React + Tailwind, and is the main interface where users can submit images and view information on identified animals. It interacts with the backend, built with Node.js, through REST APIs. The backend will process the submitted images by using Google Vision for animal identification, then is passed onto ChatGPT-4o to retrieve information about the characteristics of the animal, before returning the information. All image submissions, along with the information retrieved, is stored in a PostgreSQL instance hosted on AWS RDS for future reference.

Challenges we ran into

It is shocking to see there lack of open sourced model and data for animals, since our original plan was to train and implement our own classification model.

It is really hard for us to find a model that fulfilled all our needs.

There are some problems when integrating the endpoints with the frontend.

Accomplishments that we're proud of

We are proud that we made a social impact with modern technologies like AI, and spreading a message.

What we learned

We learned about frontend and backend development, how to use node.js and express. It is important to protect the nature.

What's next for Wildlife Watch

Improve classification granularity with model fine-tuning, Native implementation on mobile platforms, Optimize backend, Apply it in scientific research

Built With

• amazon-web-services
• node.js
• postgresql
• react
• rest
• tailwind-css