The members of Shanghai Jiao Tong University (China), has scholarly researched about identifying a criminal from their facial expression. In 19th-century scientists declaimed to inspect a criminal from their looks and now the artificial intelligence executed the research.
Two researchers from Shanghai Jiao Tong University in China namely Xiaolin Wu and Xi Zhang made a neural network for detecting an offender by their looks. For this mission they used an array of machine-vision algorithms and inspected a chain of combination among photos of offender and non- offender, with an aim of identifying the suspect with a neural network.
The researchers provided only 1856 photos to the neural network. The photos consist of all men’s photos without any facial hairs. The ages of the criminal in the photos were in between 18 to 56 and a part of them where having a criminal history.
90% of the photos were aimed to teach Artificial Intelligence for identifying among offender and non- offender and the left-over 10% were reserved for analysis use. The result of the neural network was extraordinary; the precision was 89.5 percent approx. The neural network was able to distinguish between the offender and non-offender by their facial expression.
The researchers said that “the result was quite impressive and the neural network has succeeded in its work of selecting the true culprits”. According to MIT Technology there are three features of facial expression of the criminal which the neural network remarks.
The curve of upper lip is 23 % big for criminals than for non-criminals, the space among two inner corners of the eyes is 6 % smaller than that of the non- criminals and the two outline illustrated from the tip of the nose to the corners of the mouth is 20% smaller than that of the non- criminals.
As compared to the non- criminals, criminals has big inconsistency in their facial expression. Neural networks are a replica of a human brain. We don’t know if the machine could give the real result but it will help the artificial intelligence in some case.
We cannot come to any conclusion after reading this article but we can hope that the technology can identify and give an accurate decision to the AI (artificial intelligence). Hope that the non-criminals do not suffer the decision of a machine neural network.