Computers that can identify objects seem a thing from the future. Apparently, it is more close to reality than any of us think. Birmingham Young University from Provo, US – has found a way to make computers identify objects without the need of a human helping hand.
According to Dah-Jye Lee, BUY engineer, algorithms have become so advanced that they can make a piece of software identify objects by themselves from images and even videos. Lee is the founder of this algorithm and from what he describes, it is based on the computer making decisions on its own based on the shapes identified on the images or videos analysed.
“In most cases, people are in charge of deciding what features to focus on and they then write the algorithm based off that,” said Lee, a professor of electrical and computer engineering. “With our algorithm, we give it a set of images and let the computer decide which features are important.”
Lee’s algorithm is said to learn on its own, just as a child learns to distinguish a cat from a dog. He explains that instead of teaching a child the difference between the latter, we are better off showing the two images and let the child distinguish them on his or her own. Just like a child, the algorithm has been shown four image datasets from CalTech, namely motorbikes, faces, airplanes and cars, having the algorithm output 100% accurate results on each of the datasets. However, the algorithm had a lower rate of success with human faces, being able to accurately distinguish 99.4%, but still gave a better result than other object recognition systems.
“It’s very comparable to other object recognition algorithms for accuracy, but, we don’t need humans to be involved,” Lee said. “You don’t have to reinvent the wheel each time. You just run it.”
Professor Lee mentioned that the highly complicated algorithm may be used in a variety of tasks, from detecting invasive fish species to identifying flaws in produce such as apples on a production line. However, the complexity of the algorithm can go way beyond that.