While human beings have been around for some 5000 years, our understanding of the human mind is still rudimentary. This is unlike the human physical body, which has been well researched and reasonably well understood. One classic example often cited is the difficulty in programming a robot to cross a room with the shortest path, a task most humans could do without effort. The Visual Perception Laboratory (ViPER) of the Purdue University, led by Dr Zygmunt Pizlo, has taken up a range of research projects to further enhance our understanding of the human mind. This would lead to the development of robots with more human-like skills.

Dr Zygmunt Pizlo points out that the functions of perception, thinking and memory are mind functions specific to an individual and there is no way for another person or machine to reproduce these. There is, therefore, no algorithm to mimic the way the human mind functions. It is virtually impossible to program a robot to mimic the human mind. Even in the simplest experiment of programming a robot to cross a room, the attempt is to calculate distances, photograph obstacles and then to write a program for the robot to choose the shortest path around the obstacles.

ViPER’s research team postulated that the perception of shape by the human eye is a unique attribute. Earlier, it was believed that the shape was made up of various properties of the object like color, speed, depth and motion. Dr Pizlo and his team saw that the shape of the object, as seen by the human eye, does not change even if the angle or orientation of viewing is changed. Once this uniqueness is accepted, it is possible to conduct experiments that show that the perception is unchanged regardless of the other properties. Dr Pizlo also offered a theory on how the images of the two eyes converted the 2D images seen into a 3D image. This helped prepare the mathematical model for 3D images and based on this, a new approach to machine vision has emerged. This new approach has paved the way for visual-based navigation in robotics.

The other major area of research in ViPER, related to this field, is the fact that the human mind is able to trade-off speed for accuracy. In other words, the mind comes up with a quick and approximate solution to a problem which is later refined into a more accurate solution. As new factors come into play, the mind is able to switch rapidly between the quick-approximate and the reasoned-exact responses. This too is difficult to program into a robot. Dr Pizlo’s team is working on a pyramid model of hierarchical decision making to mimic this feature of the human mind.

The work being done at the Visual Perception Lab will not just advance robotics, but also enhance our understanding of the human mind. This improved understanding will serve to bring about the next generation of robots that could more closely mimic human functions and perhaps serve to relieve humans of some of the tedious and hazardous tasks that today cannot be done by machines.

+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

About the author: Amanda Kidd is a blogger who also happens to be a tech buff. Apart from tech portal she also enjoys blogging on latest automobiles including trucks and autos. She has a weakness for cellphones and recently switched to Sony Ericsson android.

About amandakidd

Amanda Kidd is a blogger and writer who has a special inclination towards technology. Her blogs on tech involves new gadgets and technology upcoming in the market like windows smartphone.

• View all posts by amandakidd →
• Blog