Whether you are an animal or not, you cannot deny the fact that stray dogs and their increasing population is a critical issue in India. Rabies is a fatal disease and it can be transmitted to humans. Even though almost all warm-blooded animals can get and transmit this disease, dogs are the most common carrier. And today, the problem is becoming serious day by day, and abandonment and lack of sterilization are the major reasons.
In order to cope with the problem, a 12th-grade student of National Public School in Indiranagar, Bangalore, Aparna Ajit Gupte came with a solution for the vaccination attempts by civic authorities. Aparna built an AI system that uses neural networks to recognize & track stray dogs that needs vaccination. The 12th-grade student came with the idea when she saw a void in the vaccination attempts and she felt that it could be made more efficient if they had some sort of recording system.
Sign up for your weekly dose of what's up in emerging technology.
Talking about how she made that tool work, Aparna first obtained pre-trained neural networks that are capable of identifying dogs in images extracted from video recordings. Then she came with algorithms for face recognition and modified them to detect unique markings and features of the dogs. Once that is done, she just had to classify them.
Furthermore, Aparna tested her system and she successfully achieves 92 per cent of accuracy when she was identifying different dogs in a locality in Bangalore. The tool was developed in such a way by the young girl that, all you have to do is keep updating the data with images of dogs that are already being vaccinated.
That is not all, she always believes that this entire process can be automated if we use images from CCTV cameras on the streets. This would save a significant amount of time updating the images and would make the process more efficient.