Summary: New research indicates that what we perceive as visual beauty may be linked to the brain’s drive to conserve energy. Images that demand fewer neural resources and lower metabolic expense are consistently judged as more pleasing. This work combines computational modeling, human preference ratings, and brain imaging to show the visual system favors a balance between engaging input and low biological cost.
Although looking feels effortless, visual processing is metabolically expensive: the brain consumes roughly 20% of the body’s energy, and nearly half of that expenditure supports vision. The study shows that aesthetic preference aligns with energy efficiency—people favor images that stimulate the visual system sufficiently without imposing large metabolic demands.
- Energy-efficient beauty: Observers prefer images that the visual system can process using fewer neurons and lower metabolic resources.
- High cost of vision: Vision is a major consumer of brain energy, which makes efficiency an important constraint on perception.
- First-impression effect: Preferences observed here reflect rapid, automatic visual judgments rather than slower, reflective appreciation of meaning or context.
Source: PNAS Nexus
Key finding: Human attraction to certain images may partly reflect an energy-saving heuristic: the brain rewards visual inputs that hit a “sweet spot” between informative stimulation and low metabolic cost.
The researchers tested their hypothesis across three complementary approaches. First, they used an in-silico model of the visual system (VGG19) to estimate the neural resources required to represent 4,914 images of objects and natural scenes. Second, they collected aesthetic ratings from 1,118 human participants to determine which images people found most attractive. Third, they measured metabolic activity in the human visual cortex using blood oxygen level–dependent (BOLD) signals during fMRI in a smaller sample.
Across both the computational model and human data, a clear pattern emerged: images associated with lower estimated or measured metabolic cost tended to be rated as more aesthetically pleasing. Importantly, the strongest relationship appeared when the computational model had been pretrained for object and scene categorization, suggesting that experience and learned visual representations shape which stimuli are metabolically efficient and consequently preferred.
The imaging results extended the behavioral and modeling findings to the human brain. Lower BOLD responses in early visual areas (V1, V2, V4) and higher-level regions involved in scene and object processing (including fusiform and parahippocampal areas) corresponded to higher aesthetic ratings, indicating an inverse relationship between neural metabolic activity and reported liking.
Participants provided rapid, first-impression judgments rather than extended contemplative evaluations. The authors emphasize that this effect applies to fast, instinctive visual liking—not to richer aesthetic appreciation that can arise when viewers engage with an artwork’s meaning, context, or symbolism over time.
Frequently asked questions
A: The brain appears to favor images that require less energy to process. Those images hit a balance of informative stimulation with lower metabolic cost, which people find pleasant at first glance.
A: The study finds a consistent inverse relationship: images that activate fewer neurons or evoke lower metabolic signals are rated as more visually pleasing in rapid judgments.
A: This work targets rapid, automatic preferences. Deeper, reflective aesthetic experiences that depend on context and interpretation are outside the scope of these findings.
Editorial notes
- This article was edited by a Neuroscience News editor.
- The journal paper was reviewed in full by the editorial team.
- Additional context was added by staff to clarify methods and implications.
About this visual neuroscience research news
Author: Yikai Tang
Source: University of Toronto
Contact: Yikai Tang – University of Toronto
Image credit: Neuroscience News
Original research: “Less is more: Aesthetic liking is inversely related to metabolic expense by the visual system” by Yikai Tang et al., published in PNAS Nexus. This is open-access research examining the link between metabolic expense in visual processing and human aesthetic preference.
Abstract
Less is more: Aesthetic liking is inversely related to metabolic expense by the visual system
Energy efficiency is a major evolutionary driver, and prior work suggests humans use pleasure-based signals to guide adaptive behavior. The authors tested whether this energy-saving heuristic extends to aesthetic pleasure by combining an in-silico visual model (VGG19) with human behavioral ratings and fMRI measurements. They measured a proxy for metabolic cost in the model as it processed nearly 5,000 images, finding an inverse relationship between aesthetic preference and estimated metabolic cost—particularly for models pretrained on object and scene recognition. Human fMRI data showed the same inverse correlation between BOLD responses in both early and higher-level visual regions and aesthetic liking. These results propose a unified framework linking visual discomfort, processing fluency, complexity, and prototypicality: rapid aesthetic judgments may partly arise from an affective preference for lower-energy neural states.