PawsyAI

Inspiration

My app was inspired by one of my two dogs, Milky, a miniature poodle mix. We noticed that he had this small lump and didn't think much of it. My dad, who is a veterinarian, found out that it was a tumor and had to operate on it to see if it was cancerous or not. He had also told my family that the tumor was potentially cancerous, and this dog was almost about my age; since I had grown up with him, I didn't even want to think about losing him. Luckily, we were in time to operate on the tumor in time. Still, this case made me think about how it's pretty easy to fail to notice the early signs of health issues in our pet because we were only able to notice this tumor before it was too late because my dad is a veterinarian; not everybody has a veterinarian in their home who would pay attention to this kind of thing. I have two dogs, and sometimes when one of them has some possible health problem, such as a hot spot, I don't know how to find out what it is, and I have to wait for my dad to come home from work or from his business trip, where he can sometimes take a long time since he is very busy. I wanted to develop a tool that would contribute to finding any sort of problems much sooner rather than later, when it might already be too late for someone who owns a pet and doesn't have immediate access to a veterinarian at all times.

What it does

"Using Convolutional Neural Networks to Detect and Prevent Potential Health Issues in Dogs By Using a Mobile Application." This mobile app lets the user take a picture of their dog or upload a picture for it to be analyzed by our machine learning model in search of potential health issues like skin, eye, mouth, and ear abnormalities. Through a TensorFlow machine learning model that we developed, the app will tell the user if a potential health issue is detected. Using the built-in AI chatbot, the app will help advise the user based on the prediction they received through the machine learning model of their dog's potential health issue. Through this app, it is very easy for pet owners to ensure that their pet is healthy without needing to immediately contact a veterinarian every time they notice a potential health issue in their pet. This app was made with React Native, TensorFlow, and Gemini API.

How we built it

We built the app using React Native, used TensorFlow for the machine learning model, and used Gemini for the chatbot.

Challenges we ran into

Getting the TensorFlow model integrated into the app took a lot of time and was super stressful, but we eventually figured it out.

Accomplishments that we're proud of

The model is decently accurate, and whenever I tried it on my dogs, it was correct. My dad is a veterinarian, but even though I want to pursue computer science when I grow up, I am going to use this app as a way to connect with my dad.

What we learned

I learned how to use TensorFlow in apps and plan on making a lot of them for hackathons as it is a cool way to make an impressive project.

What's next for PawsyAI

I am planning on writing a research paper on this app and submitting it to ISEF. I have huge plans for this app and want to eventually publish it to the app store. I want to add more complex features to this app to be a one-stop shop for all pet owners.