PupTrack

Inspiration

One of our own member's worry about his puppy inspired us to create this project, so he could keep an eye on him.

What it does

Our app essentially monitors your dog(s) and determines their mood/emotional state based on their sound and body language, and optionally notifies the owner about any changes in it. Specifically, if the dog becomes agitated for any reasons, manages to escape wherever they are supposed to be, or if they fall asleep or wake up.

How we built it

We built the behavioral detection using OpenCV and TensorFlow with a publicly available neural network. The notification system utilizes the Twilio API to notify owners via SMS. The app's user interface was created using JavaScript, CSS, and HTML.

Challenges we ran into

We found it difficult to identify the emotional state of the dog using only a camera feed. Designing and writing a clean and efficient UI that worked with both desktop and mobile platforms was also challenging.

Accomplishments that we're proud of

Our largest achievement was determining whether the dog was agitated, sleeping, or had just escaped using computer vision. We are also very proud of our UI design.

What we learned

We learned some more about utilizing computer vision and neural networks.

What's next for PupTrack

KittyTrack, possibly Improving the detection, so it is more useful for our team member

Built With

• html
• javascript
• opencv
• python
• sms
• tensorflow
• twilio