Release date: 2016-09-23

IBM and the Massachusetts Institute of Technology (MIT) announced on the 20th that they will jointly create the "Braining Multimedia Machine Comprehension Lab (BM3C)", which aims to make artificial intelligence look and listen like people.

When people watch a baseball game, it is easy to distinguish the pitcher from the pitching area, and even predict whether the next pitching is a curved ball or a fastball. However, there is still no machine that can understand these very simple for humans. Mission, IBM and MIT scientists hope to change this.

According to the US "Christian Science Monitor" reported on the 20th, they will jointly create a BM3C to carry out related research, so that the machine can see, hear and interpret what it sees like a human being, and this project will last for many years.

The lab is the latest in a number of collaborative programs that IBM is pushing to advance the development of artificial intelligence. At present, many research teams are engaged in research that allows machines to "think like humans and even surpass humans in some ways."

Previously, other research teams at MIT have made advances in computer prediction and understanding. For example, in June of this year, scientists at the Computer Science and Artificial Intelligence Laboratory (CSAIL) announced that they had taught them how to say hello by shaking hands or hugs by showing the machine a 600-hour video and TV show from YouTube. The accuracy rate is 43%; the accuracy rate of human experimenters is 71%. In 2015, it was reported that two other researchers at MIT created the Data Science Machine (DSM), which can uncover data patterns and pick out relevant data points without help.

Still, teaching machines to look and listen like people is still not done. Devin Kidway, TechCrunch, a technology blogger, explains that the problem is that human vision and voice commands span multiple cognitive rules: humans first need to discover objects, then establish relationships between them, and intuitively obtain physical rules, and so on. These things are especially good at our brains.

Since human thinking has mastered these skills, the researchers plan to model machine vision on a virtual neural network based on real things. They expect that machine vision systems can have a huge impact on industries such as education, entertainment, and health.

Source: Technology Daily

Travel Testing Kit

Characteristics disease virus antibody test kit Wondfo Antibody Rapid Test Kit No cross reaction with common respiratory viruses Rich sample type suitable for medical institutions Easy to operate safe

Travel Testing Kit

NINGBO AUTRENDS INTERNATIONAL TRADE CO., LTD , https://www.metests.com