Artificial intelligence will soon understand sight and sound like we do

When you see or hear something happen, you can instantly describe it. For example: “The black and white dog caught a tennis ball thrown by the man wearing a red hat.” It doesn’t take much effort for a human to describe, but when it comes to computers, such a task is not so simple. Thanks to a collaboration between IBM and MIT, that’s about to change. 

The computer hardware company and private university have teamed up to support the development of machine vision using insights from brain and cognitive research. The new IBM-MIT Laboratory for Brain-inspired Multimedia Machine Comprehension, also known as BM3C, is a multi-year collaboration between the two organizations that will be digging deeper into the problem of computer vision and audition.

BM3C will look to find answers to numerous technical challenges around pattern recognition and prediction in machine vision — which are impossible tasks for machines to complete alone. The statement proposes an example in which a human watches a short video of a real-world event and can easily produce a description of the clip, as well as assess the likelihood of subsequent events — all of which are impossible for a machine to accomplish.

Beginning this month, BM3C will bring together leading brain, cognitive, and computer researchers to conduct further study into unsupervised machine understanding of audio-visual data streams. The team’s goal is that the cross-discipline approach will lead to advances in machine vision, and for use across many industries such as healthcare, education, and entertainment.

The BM3C project will be led by Professor James DiCarlo, who currently leads MIT’s Brain and Cognitive Sciences (BCS) department. DiCarlo will be supported by a team of faculty members, researchers, and graduate students from both the BCS department and the MIT Computer Science and Artificial Intelligence Lab (CSAIL). This team of academics will work alongside IBM scientists and engineers, and the IBM Watson platform.

Source: PR Newswire