Videos with more movement can feel longer
Adults estimated video durations. A program that added up image changes reproduced some of those trends, but not every estimate.
The key idea
Felt duration varied with what happened on screen, and a program that added up image changes reproduced part of that pattern.
The study, in plain language
The clock says the same amount of time, but a walk through a busy city and a quiet office scene can feel different in length. To study this, the team showed natural videos to young adults, asked them to estimate each duration, and recorded where they looked.
Duration estimates varied with the visual content of the scenes. A program that accumulated image changes reproduced some broad features of those responses. Interestingly, it had been fitted to clock-measured durations, not people’s estimates, yet similarities appeared between the two patterns.
What they found
- Scenes with more visual activity and quieter scenes felt different in duration.
- A program that accumulated image changes reproduced some trends in human estimates.
- The program was fitted to clock-measured durations, not participants’ responses.
Something to think about
The next time a video feels unusually long, compare that feeling with what the clock says and notice how much the image changes.
How they studied it
Fifty-five young adults watched videos, estimated their durations, and had their gaze recorded. Videos lasted from one to 64 seconds, and the number of trials varied between participants. The team compared these responses with a model that accumulated changes in the activity of an image-classification network. The model was fitted to clock-measured durations, not human responses, so its similarity to them does not show that the brain uses exactly this mechanism or that perceived time depends only on vision.
What the study cannot show
- The sample was young and university-based, and the task was limited to visual videos.
- The program is a computational hypothesis. The study does not show that the brain uses this algorithm or that time perception depends only on vision.
The original study
Activity in perceptual classification networks as a basis for human subjective time perception
Source under CC BY 4.0. English adaptation and explanation by PaperFlow, with AI assistance. The text has been reorganized for a general audience; it is not the complete paper and does not imply endorsement by its authors.
One study alone does not represent a field's consensus. Educational content does not replace professional advice. Report a possible error.
