When the children were shown a new object, such as an alphabet letter, a photo or an element of the world ,they just need some time to identify it accurately. While robots need to be trained hundreds and thousands of times to identify not only an object, but also to identify it from different angles.
To solve the problem,scientists have developed an algorithm which allows the robots to learn visually in the same way as humans do. This has allowed them not only to identify an object by its shape, but also to attract it for themselves, just like young children do when they are being taught.
Professor Joshua Tenenbaum, a scientist at the Institute of Technology in Massachusetts, who was also involved in the study, said: “Before Babies go in the garden, they learn to recognize new concepts of an example only and can even imagine other examples, they have not seen. We are still far away from the construction of robots smart as a human child, but this is the first time we had one, which is able to learn. ”
This algorithm has also been established with the cooperation of Dr. Brenden Lake, a scientist at New York University. The robot was tested in more than 1,600 types of handwritten symbols in 50 different alphabets or codes. They included Sanskrit, Tibetan, Gujarati, Glagolitic, as well as imaginary letters shown TV series. The program was able to consistently reproduce these symbols, which you shafaqen only once.
No comments:
Post a Comment