This Girl To Develop An AI System To Detect & Follow Street Dogs

Jyotis - Jun 03, 2019

That’s exactly what a Bengaluru student thought when developing artificial intelligence (AI) to detect and keep track of streets dogs for vaccination.

One of the existing problems in India hails from street dogs, and it’s the true animal lovers can’t deny. There are a lot of reasons why the number of street dogs has been increasing in many cities. They include abandonment, a shortage of rational sterilization programs, and careless breeding.

Stray-Dogs-Ai — A new AI has been created to recognize and keep track of the stray dogs.

In some unavoidable circumstances, owners have no choice but to throw away their dogs or other pets. However, they should at least keep these animals healthy or ensure they won’t harm the citizens nearby. The simplest solution is to inject vaccine for them. That’s exactly what a Bengaluru student thought when developing artificial intelligence (AI) to detect and keep track of streets dogs for vaccination.

The student’s name is Aparna Ajit Gupte who is in grade 12 at National Public School, Indiranagar. This idea came to her when Gupte read about a series of human deaths on news articles and a large part of these accidents are caused by rabid street dogs. Although civic authorities have made great attempts on vaccination, she realized that it would be better if a recording system is available. That’s why the Bengaluru student decided to create such an AI.

AI-Dogs — The accuracy of this AI reached up to 92%.

In the beginning, Gupte developed neural networks that are pre-trained and can detect dogs via images taken from video recordings. And next, she applied an algorithm to recognize dog faces and changed to discover features and markings only available in street dogs.

It’s worth mentioning that the accuracy of her initiative reached up to 92% when considering the ability to identify the various dogs in an area. As such, users just need to add images of the vaccinated dogs to update the data and leverage the AI. In the upcoming time, users will easily recognize the dogs that haven’t been tackled.

In addition, Gupte also suggested footage extracted from CCTV cameras on the street should be used to track the dogs and automate the AI system. Next time, she is going to keep track of the number of dogs in her area with a hope to keep the dataset updated via new images.