+1-518-621-2074 | US-Canada Toll Free Contact Us

Has AI Bridged the Human - Computer Gap Through Visual Recognition?

about us

Published on : Mar 25, 2019

Researchers are focusing on making computers think like humans so as to reduce the manual work. They have achieved this to a considerable extent by leveraging the power of artificial intelligence. However, although computers offer a greater accuracy than humans do, technology always has certain loopholes. There is a perpetual threat of data leakage and hackers gaining access to important files.

New Research to Test Visual Abilities of Computers against Humans

Chaz Firestone, an assistant professor in Johns Hopkins' Department is working on a project to make humans think like computers. Firestone along with Zhenglong Zhou, a Johns Hopkins senior majoring in cognitive science asked people to think like a machine. They showed participants specific images that computers failed to identify. With this, the researchers gauged the recognition capabilities of people as well as the computers. . Further, they asked people to choose between the answer provided by the computer and a random answer. The results suggested that people’s answers matched with the computer 75 percent of the times.

Furthermore, researchers provided a choice between a computer’s favorite conclusion and its second-best guess. The findings suggested that people agreed with the computer’s favorite solutions 91 percent of the time. An overwhelming number of people agreeing with the computer’s conclusions suggests there is a major shortfall in the technology. Firestone said that humans are most likely to agree with a computer when made to experience the same circumstance. Although this is a problem for artificial intelligence, it is not something we can’t overcome. Moreover, the research proves that computers are not as different from humans as we perceive them to be.

Recent development in regards to neural networks have given machines the ability to mimic the human brain. This has led to greater technological advancements such as self-driving vehicles and facial recognition programs. In the future, advances in artificial intelligence will enhance the visual abilities of the computer and tackle the probable safety-threats.