Getting a headstart on predictions

The human ability to predict the future actions of other humans and objects is a natural reaction while watching a movie or drama series. However, teaching computers to do so is not as simple.

"Action anticipation models should be able to account for future uncertainties. For the same observations, there are numerous plausible futures and these models should be able to predict all of them accurately,” said Basura Fernando, a scientist at A*STAR’s Institute of High Performance Computing (IHPC) and recipient of an A*STAR scholarship.

Dr Samitha Herath, Research Fellow at Monash University in Australia, and Dr Basura have developed a framework for an action anticipation model that correlates past observations with future actions. They then used these correlations to extract predictions from current observations. 

Read the full article published on .