Focusing on low-latency, low power, compact adaptive vision systems through back-to-back integration of sensing and vision processing

Funded by the Office of Naval Research (ONR), the Carnegie Mellon University MURI Program  works on vision systems which  may lead to a wide variety of applications including  augmentation of the human sensory system for "natural" information delivery.  The project explores:

• SMART OPTICS based on the Acousto-Optic Tunable Filter (AOTF) technology;


• VLSI COMPUTATIONAL SENSORS which integrate raw sensing with computation in a single chip;


• VISULA SALIENCY DETECTION based on connectionist as well as model-based computation;


• LEARNING AND RECOGNITION of objects via automatic visual model acquisition;


• ADAPTATION  to problem domains by using genetic programming and artificial evolution;


• INTEGRATION of the above areas into a real-time vision systems for autonomous road navigation, helicopter flying and robotic soccer playing.

Our multidisciplinary team  builds upon present and past research successes  in several CMU departments, including Carnegie Mellon Research Institute (CMRI), School of Computer Science (SCS), The Robotics Institute (RI), and Electrical and Computer Engineering Department (ECE).