Internet company Google and social network company Facebook have been involved in an algorithm war to come up with the perfect facial detection system. Now, California headquartered Google claims that it has developed new FaceNet system, which is practically perfect. According to the company, the FaceNet system gets the right person more than 99% of the time.

According to researchers of the Internet Company, the new system that they had developed is the most accurate technology available at present to recognize human faces. The company has claimed the technology with title 'FaceNet: A Unified Embedding for Face Recognition and Clustering' in a paper. As per the company's claim, the new system achieved about 100% accuracy rates on the facial recognition dataset Labeled Faces in the Wild.

To measure the accuracy of the system, the researchers used more than 13,000 face images from the Internet. About the new system, the researchers said, "Our method uses a deep convolutional network trained to directly optimize the embedding itself, rather than an intermediate bottleneck layer as in previous deep learning approaches. To train, we use triplets of roughly aligned matching / non-matching face patches generated using a novel online triplet mining method".

They said the approach has much greater representational efficiency. The system achieved a new record accuracy of 99.63% on the widely used Labeled Faces in the Wild (LFW) dataset. On YouTube Faces DB, the system's results were 95.12% accurate, as per the researchers.

In 2014, a group of researchers from China had claimed that they developed a system that achieved better than 99% accuracy. In June last year, researchers from Facebook said humans analyzing pictures in Labeled Faces dataset could only achieve about 97% accuracy.