The primary focus of my research is human object recognition.  In particular, I am interested in discovering how objects are represented in memory.  For example, it may be the case that we are able to recognize a chair as a “chair” because we have seen that chair before, and we have a “template” or picture of that chair stored in memory.  When we later see that chair, our brains find the matching picture in memory, and we recognize the object as a chair.  An alternative view suggests that rather than a picture of a chair stored in memory, we have a general description of a chair that can be applied to many different chairs, each with slight differences in shape and design.  By storing objects in this manner, we can recognize chairs that we’ve never seen before, and we can use the same label, “chair,” for chairs of many different shapes and sizes.

Through a better understanding of how objects are represented in memory, and consequently recognized, we can better predict situations in which recognition is likely to fail, and we can potentially advance progress in developing artificial vision systems.

Recent Research Presentations (Posters)

Recent Research Projects:

Identification of specific visual primitives used in basic level object recognition

Although to the observer the process is effortless, at all times as individuals interact with their visual environments they are identifying the objects with which they are surrounded.  Despite the apparent ease of accomplishing this task, the neurological and psychological processes involved are extraordinarily complex.  Part of the complexity stems from the problem of grouping various objects that are not identical to one another into the same general category.  For example, despite the fact that two chairs might look significantly different from one another, an observer typically has no trouble whatsoever identifying both objects as chairs.

In order to classify and recognize different exemplars of the same object as belonging to the same object class, our visual systems employ the use of structural descriptions--general descriptions of objects which include the parts of the objects and the relations of those parts to one another.  For example, a coffee mug might be classified as a curved cylinder to the side of a cylinder.  By specifying only the parts and the relative relationships between those parts rather than the exact size of the parts or the exact relationships between the parts, many different sizes and designs of coffee mugs can be quickly and easily grouped into the general recognition category of “coffee mugs.”  In comparison, a pail might be classified as a curved cylinder above a cylinder, but despite the fact that the pails and coffee mugs share the same basic parts (a cylinder and a curved cylinder), we can differentiate between the two because the relative relationships between the parts of the objects (to the side of, vs. above) differ between the two objects.

I have an ongoing line of research designed to identify the exact nature of the basic primitives (or parts) that we use as the foundation for recognizing any object.  Very much like we use an alphabet of 26 letters to create many thousands of words, it appears to be the case that our visual systems use a basic “alphabet” of shapes to recognize many thousands of objects.  Just as we are able to recognize words printed in different fonts as being the same word given that the letters are arranged in the correct order, we are able to recognize objects with small variations in the design as being the same object, given that the objects have the same parts and relationships between the parts.

As part of this research I have had the opportunity to work with a patient afflicted with a disorder known as prosopagnosia, and this opportunity has allowed me to address my research questions from multiple perspectives.  Prosopagnosia is the inability to visually recognize faces, while retaining the ability to recognize individuals via voice, gait, body size, or other distinctive features.  By identifying the recognition tasks that cannot be accomplished by an individual with selective impairments in object recognition, I have been able to more accurately explain and predict the recognition abilities of individuals with normally functioning visual systems.

Perception, language, & culture

Another project that I have been working on over the course of the last few years involves the influence of environmental effects on visual perception.  In particular, I am interested in whether exposure to certain visual stimuli, such as color, can influence an individual’s subsequent ability to physically perceive various colors.  For example, if an adult had never seen the color purple, could that adult visually discriminate between two slightly different shades of purple as accurately as an adult that had been exposed to the color purple from a very young age?  As part of this project, I have been collecting data on color categorization patterns from participants living in continents other than North America.

If you are interested in learning more about object recognition research or independent psych research credit, please contact me at obrien.alex@uwlax.edu