Sep 20, 2019

MoooFarm achieves 95.7% accuracy in facial recognition software for cows

MoooFarm, an India-based startup, is building facial recognition software for cows that will help farmers identify cattle using their smartphone cameras.

As countries such as the U.S. debate the ethics of facial recognition and artificial intelligence in terms of their privacy impact on humans, an entire world of software developers are pushing technology forward.

The MoooFarm team recently received a $30,000 award from the World Bank for their facial recognition technology that achieved 95.7% accuracy in identifying individual cows. MoooFarm previously received a $15,000 grant from Microsoft to tackle mastitis, a disease that negatively affects cows’ udders, using image labeling technology to analyze photos of cows to detect instances of the disease.

Like humans, cows have defining facial features that can be used to distinguish them from each other. Farmers in India can take pictures of their cows to build a directory of their livestock, minimizing stray cattle and preventing insurance fraud. MoooFarm’s software works with basic smartphones, making the tool easy to share with remote or rural farmers.

Most importantly, much of that technology is becoming increasingly accessible. For example, over the last few months, several open source repositories have sprung up that leverage advanced machine learning to create convincing fake images and videos. Small teams and independent developers have access to both immensely powerful artificial intelligence algorithms and widespread access to cloud-based platforms.

Advancements in widely accessible clouds bring with them greater access to artificial intelligence and machine learning, even in rural India. After cows, who will lose their digital anonymity next?

Want to get more of these in your inbox?

Subscribe for weekly updates from the Software team.