MIDS Capstone Project Fall 2021

AI Dog Sitter

Pets play an important part in their owners’ lives, and caring for a sick pet is a costly and  distressing experience. Early diagnosis and timely treatments of a pet’s illness can prevent further deteriorations, and understanding a pet’s typical behaviors is important for spotting early signs of illnesses. 

AI Dog Sitter is a versatile, low cost monitoring tool for tracking activity patterns in dogs, the most popular pet in the world. It uses deep learning algorithms trained on an open dataset of annotated dog videos to identify a dog’s activities captured in a video stream. Owners and veterinarians get access to a dashboard displaying analyses of recent activities as well as comparisons to past averages, so they can gain better insights into the dog’s current health. 

As a proof of concept, a web-based Minimum Viable Product is deployed online. It processes uploaded video clips to produce a detailed dashboard displaying the activities captured in the clip.



AIDogSitter demo

AIDogSitter demo

If you require video captions for accessibility and this video does not have captions, click here to request video captioning.

Last updated:

December 10, 2021