Smooth pursuit conduct of humanoid with visual image swap

The purpose of smooth pursuit (SM) conduct by humanoids is to keep the moving object of interest at the foveal vision of both eyes. To keep target in foveal vision is an inherit property of humans. Visual neurons are able to procure relaxed information compared to other case. Although various models are developed to perform smooth pursuit efficiently but the latency in movement is still an existing problem to be addressed and these latencies in eye ball movements are still a non-trivial challenge. Previous work addresses visual encoding and eye-control movements while considering smooth pursuit a behavior of eye balls. Our model accounts for real time development of visual image swap phenomenon with joint functionality of visual encoding and controlling for efficient processing. Visual image swap is process of exchanging images from one eye to another. The purpose of exchanging images is to improve processing speed and as a result to shrink latency. Smooth pursuit approach is enhanced from eyes to mutual synergy of eyes and head movements. We modeled and tested this approach on a real robot. Experiments conducted on the robot demonstrated that by executing visual encoding and controller operation jointly along with visual image swap resulted in improved gain (ratio of target velocity to organ velocity) and decreased the delay not just for eyes, for head also. Eyes head movements of the robot are similar in characteristics to human smooth pursuit behavior.