Scenes, Vision and Pigeons. . . Oh My!

Adam Larson, UF assistant professor of psychology, may not be in Kansas anymore. That didn’t stop the online newsroom at Kansas State University (KSU) from publishing a lengthy article detailing his contribution to research on how people comprehend the meaning of a scene. Namely, how do we recognize that an image is of a beach, city or forest? Working with KSU psychology professors Lester Loschky, Kim Kirkpatrick and others, Larson contributed to an article, “The Spatiotemporal Dynamics of Scene Gist Recognition,” published in the Journal of Experimental Psychology: Human Perception and Performance.
Larson, a former student of Loschky, and two other researchers, wanted to know how humans identify a scene – specifically what portion of the visual field is used first. Images were manipulated so that scene information was presented to either central or peripheral vision. They discovered that when scenes were processed for very short periods of time (1/10^th of a second), people were better at recognizing scenes using central vision. However, when scenes were processed for longer periods of time (greater than 1/10^th of a second), there was no difference in scene recognition between central or peripheral vision.
According to Larson, this can be explained by how we distribute our attention across the visual field. Namely, people place their attention in central vision first, before they begin to spread their attention to peripheral vision. He added that this was evident in follow-up studies showing that when people were trained to focus their attention in the center of vision, then early processing of scenes produced better recognition in central vision. Conversely, if trained to focus their attention in the visual periphery, then scene recognition was not different between central and peripheral vision. This shows that where we attend in a scene changes over time and that attention plays a critical role in how we recognize the meaning of a scene.
Researcher Kim Kirkpatrick even worked with pigeons, comparing their scene recognition results with results of tests involving humans. She discovered that pigeons can recognize scenes like humans, but it takes the birds about 10 times longer to do so.
Larson said this type of research could improve driver safety because it contributes to knowledge about how people discern and attend to a scene. He added there is also value for the field of robotics where scientists are working on developing “human-like” vision with computers.
The Path to Findlay
Larson is in his second year at UF, teaching two sections of General Psychology; a class in Sensation and Perception and one in Cognitive Psychology. He encourages student involvement in research.
Growing up in Iowa, Larson received his bachelor’s degree from Iowa State University. He chose Kansas State University for his master’s and doctoral studies because, “They had very knowledgeable faculty and the resources needed to become an expert in visual cognition.”
He’s now interested in studying how people comprehend and pay attention to film. Some of the preliminary research has suggested that regardless of how well people comprehend a film, everyone appears to be looking at the same place, at the same time. This suggests that film directors can create a shot sequence that captures the viewer’s attention. Eventually, he’d like to work with filmmakers, tracking the audience’s eyes on the screen.
“I have a colleague who went to Los Angeles to work with people in the film industry. It’s valuable to them to know how people are looking at the screen and what they are able to understand at that time.”

---