Accomplishments of the Neural Computation and Adaptive Perception Program
 

• Program members in CIFAR’s NCAP program have succeeded in creating a synthetic neural network that can recognize familiar objects, even when significantly distorted or observed from a new viewpoint. This innovation results from research into how the human brain learns, takes in visual information and filters out irrelevant stimuli. Some practical applications of this discovery include tools that recognize hand-written postal codes and the ability to translate written material into computer text.
• CIFAR researchers have also been successful in modeling human motion and extracting it from video sequences. These findings support efforts to create new visual systems that can be applied in areas such as artificial intelligence and sophisticated security surveillance systems.
• The series of neural transformations that take place in the human brain when recognizing a face are complex. New, more precise methods for determining this specific sequence of events have been developed and successfully applied by NCAP researchers. Practical applications may include computer-based security systems that are able to detect and identify human faces.