University of California, Santa Barbara (UCSB) researchers have demonstrated a neural circuit of about 100 artificial synapses that is able to perform a simple version of image classification. The researchers say with more time and development, the technology could be expanded to approach the functionality of the human brain, which has one quadrillion synaptic connections. In the demonstration, the artificial neural network circuit was able to successfully classify three letters by their images, with each letter stylized in different ways or obstructed with noise. “While the circuit was very small compared to practical networks, it is big enough to prove the concept of practicality,” says UCSB researcher Farnood Merrikh-Bayat. The technology relies on the memristor, an electronic component whose resistance changes based on the direction of the flow of the electrical charge. Memristor operation depends on ionic movement, similar to the way human neural cells generate neural electrical signals. UCSB professor Dmitri Strukov notes the ionic memory mechanism produces several advantages over electron-based memories, which makes it an attractive solution for artificial neural network implementation. However, in order to be able to approach the functionality of the human brain, many more memristors would be required to build more-complex neural networks.
More info here: The UCSB Current (05/11/15) Sonia Fernandez