Why images may appear closer, or farther, than they actually are

Summary: Boundary contraction may be as common as boundary extension, and whether something appears zoomed in or out depends on the properties of the image.

Source: University of Chicago

When people remember images, they fill in the edges with details they didn’t actually see. That’s the idea behind the boundary extension, a term which has become widely accepted in psychology classes, textbooks and test-prep flashcards.

But what if the concept isn’t quite accurate?

A University of Chicago psychologist has discovered new evidence that challenges the decades-old understanding of the memory error as a universal phenomenon. Published in the journal Current Biology, the study proposes that boundary contraction may be just as common as boundary extension—and that whether something appears zoomed in or out depends on the properties of the image itself.

“In a way, we’re debunking this very strong claim that has been made in psychology over the last 30 years,” said Asst. Prof. Wilma Bainbridge, the study’s lead author and an expert on the perception and memorability of images.

The finding is important, she added, because boundary extension has been used to make other claims about the nature of the brain, such as the function of the hippocampus.

Bainbridge co-authored the study with Chris Baker, a principal investigator at the National Institute of Mental Health. Testing 2,000 participants, they found that although images of objects caused boundary extension, images of full scenes were more likely to produce boundary contraction. That is, a person may see a close-up photo of an apple and fill in details that were not actually present. But if they see a football field, they may be more likely to remove details—zooming in, or contracting, the actual image.

In a previous study, Bainbridge and Baker showed participants various images and asked them to draw copies. They were “perplexed” when boundary extension did not occur as often as they had expected.

This shows images from the study

The study relied on two databases, Google Open Images (SOI) and the Scene Understanding Database (SUN), which categorized images using object- and scene-oriented words. The image is credited to Wilma Bainbridge.

To further investigate those results, they conducted an online experiment using a broad set of 1,000 images and 2,000 participants. Participants would see an image, see a scrambled image and then see the original image again.

Even though the final image was identical to the first, the researchers found that people would indicate it being farther or closer according to its visual properties (object-based vs. scene-based).

Bainbridge said the results highlight the need for psychologists to revisit even long-held assumptions, as well as the potential pitfalls of drawing larger inferences from limited data sets.

Past replications of boundary extension, she suggested, could have been skewed in part by narrow data sets that repeated the use of certain image types.

“Anecdotally, I’ve spoken with many people who have thought about looking at boundary extension—but then they aren’t able to replicate the effects, so they give up and they set aside the data,” she said.

About this neuroscience research article

Source:
University of Chicago
Media Contacts:
Jack Wang – University of Chicago
Image Source:
The image is credited to Wilma Bainbridge.

Original Research: Closed access
“Boundaries extend and contract in scene memory depending on image properties”. Bainbridge and Baker.
Current Biology doi:10.1016/j.cub.2019.12.004.

Abstract

Boundaries extend and contract in scene memory depending on image properties

Highlights
• Boundary contraction is as common as boundary extension for naturalistic scene images
• Boundary transformation direction is highly predictable from basic image properties
• Boundary transformations occur during recognition, recall, and minimal memory load
• These results dispute boundary extension as a universal phenomenon of scene memory

Summary
Boundary extension, a memory distortion in which observers consistently recall a scene with visual information beyond its boundaries, is widely accepted across the psychological sciences as a phenomenon revealing fundamental insight into memory representations, robust across paradigms and age groups. This phenomenon has been taken to suggest that the mental representation of a scene consists of an intermingling of sensory information and a schema that extrapolates the views of a presented scene, and it has been used to provide evidence for the role of the neocortex and hippocampus in the schematization of scenes during memory. However, the study of boundary extension has typically focused on object-oriented images that are not representative of our visuospatial world. Here, using a broad set of 1,000 images tested on 2,000 participants in a rapid recognition task, we discover “boundary contraction” as an equally robust phenomenon. Further, image composition largely drives whether extension or contraction is observed—although object-oriented images cause more boundary extension, scene-oriented images cause more boundary contraction. Finally, these effects also occur during drawing tasks, including a task with minimal memory load—when participants copy an image during viewing. Collectively, these results show that boundary extension is not a universal phenomenon and put into question the assumption that scene memory automatically combines visual information with additional context derived from internal schema. Instead, our memory for a scene may be largely driven by its visual composition, with a tendency to extend or contract the boundaries equally likely.

Feel Free To Share This Visual Neuroscience News.
Join our Newsletter
I agree to have my personal information transferred to AWeber for Neuroscience Newsletter ( more information )
Sign up to receive the latest neuroscience headlines and summaries sent to your email daily from NeuroscienceNews.com
We hate spam and only use your email to contact you about newsletters. We do not sell email addresses. You can cancel your subscription any time.