Do Vision Encoders Exhibit Human-like Color Thresholds?

arXiv:2607.16540 · cs.CV · Submitted 2026-07-17 · Read on arXiv

Listen

Radio episode about this paper

Transcript

Introduction to the show: ident: AI Radio. Generated commentary on the latest Artificial Intelligence papers.

Tom: We're looking at the paper: "Do Vision Encoders Exhibit Human-like Color Thresholds?".

Jane: Here's the gist of the paper. Understanding and characterizing human color perception is a longstanding research goal <ref:2607.16540#pg2> One of the most traditional approaches is looking for the human color discrimination thresholds, the minimum chromatic differences perceptible to human observers <ref:2607.16540#pg2>.

Tom: In this work, we present a large-scale exploratory study probing the chromatic sensitivity of more than 50 pretrained vision encoders, including convolutional networks and vision transformers, against human discrimination thresholds <ref:2607.16540#pg2>.

Jane: Our analysis reveals generally weak alignment between model representations and human perceptual thresholds across all model families, with the best mIoU.

Tom: Pruning redundant reasoning tokens The study systematically evaluates the chromatic representation geometry of deep vision encoders against human discrimination behavior <ref:2607.16540#pg2>.

Paper summary: Tom: So, wrapping up this discussion on "Do Vision Encoders Exhibit Human-like Color Thresholds?", the paper confirms that while deep vision encoders develop structured and anisotropic chromatic similarity geometries in their latent spaces, these representations only partially align with human perceptual models.

Jane: The alignment varies systematically across chromaticity regions and saturation levels, and it depends strongly on the training paradigm; self-supervised and supervised models outperformed language-supervised and generative approaches in terms of this alignment.

Lu: The overall message is that even the best performing models remain far from human-level chromatic discrimination when tested against these specific human perceptual thresholds.

Meng: It means we can't just assume high performance on a benchmark automatically translates to accurate color perception, especially since the alignment depends so heavily on what kind of training signal you feed the AI.

Tom: Exactly. The implication is that for any application where precise chromatic discrimination is critical, we need to consider both the model architecture and the specific training objective alongside just looking at raw performance metrics.

Jane: It's a caution to keep in mind as we move forward with these large foundation models, reminding us that internal feature representations don't always map perfectly onto established human psychophysical data.

Conclusion: Tom: So we’re done with looking at the mechanics of how these vision encoders are trained, and now we’re just looking at what this whole study actually means for color science.

Jane: The title is pretty direct, right? "Do Vision Encoders Exhibit Human-like Color Thresholds?" It’s asking if the AI's internal understanding of color matches what a person can actually see when they look at two colors side by side.

Lu: Basically, they took fifty different types of vision models, from older convolutional networks to the newest vision transformers, and compared how well their internal representations line up with actual human color discrimination thresholds.

Meng: And the main finding is that the alignment isn't great across the board; it’s weak generally. They found that while these models develop some structured geometry in their latent space, they don't perfectly mirror how humans perceive those differences.

Tom: So, if we think about this for a practical application—say, an autonomous vehicle needing to detect subtle color changes—this study suggests we can't just trust the raw output of these massive models when it comes to precise color discrimination.

Jane: Exactly. It shows that even the best performing models still have a gap between their internal math and real human perception, and this gap changes depending on the specific shade of blue you’re looking at or how saturated it is.

Lu: The study points out that the alignment varies wildly based on the training signal; self-supervised methods actually kept their color relationships better than language-supervised ones did.

Meng: That’s important because if we are deploying these models, choosing the right training method matters more than just picking the biggest model size.

Tom: It really highlights that achieving high scores in a test doesn't automatically mean you’ve captured human sensory reality; there's still a distance to cover.

Jane: And it leaves us wondering what that distance looks like when we move beyond simple flat color patches and look at complex, real-world textures. That’s where the next part of this line gets interesting.

Computer Vision Center, Spain · Universitat Autonoma de Barcelona, Spain · Image Processing Lab, Universitat de Valencia, Spain · Arab Academy for Science, Technology, and Maritime Transport

cs.CV

Submitted: 2026-07-17

Updated: 2026-11-02

License: http://creativecommons.org/licenses/by/4.0/

Importance score: 77/100

The gist: Understanding and characterizing human color perception is a longstanding research goal <ref:2607.16540#pg2> One of the most traditional approaches is looking for the human color discrimination…

Key concepts

Human Color Discrimination Thresholds
These are the minimum differences in color that a human observer can perceive as distinct. The study uses these thresholds as a benchmark to see how well computer vision models understand what humans can see in terms of color separation.
Model Representations Geometry
This refers to the shape and structure of the colors learned by deep vision encoders inside their internal feature spaces. The researchers analyzed whether these learned color patterns resemble the natural, anisotropic shapes (like ellipses) that describe how humans perceive color differences.
mIoU (Mean Intersection over Union)
This is a metric used to measure the overlap between the chromatic discrimination thresholds derived from a specific vision model and the actual human discrimination ellipses. A higher mIoU indicates better alignment between what the AI model thinks about color and what humans actually perceive.

Terminology

Summary

Understanding and characterizing human color perception is a longstanding research goal <ref:2607. p. 0

Where the paper says it

Do Vision Encoders Exhibit Human-like Color Thresholds? Engy Ehab 1,2,4 , Pablo Hernandez-C ´ amara ´ 3 , Nahla Belal 4 , Jes ´us Malo 3 , Javier Vazquez-Corral, 1,2 , Alexandra Gomez-Villa 1,2 1 Computer Vision Center, Spain 2 Universitat Autonoma de Barcelona, Spain ` 3 Image Processing Lab, Universitat de Valencia, Spain ` 4 Arab Academy for Science, Technology, and Maritime Transport, Egypt Abstract Understanding and characterizing human color perception is a longstanding research goal.

16540#pg2> One of the most traditional approaches is looking for the human color discrimination thresholds, the minimum chromatic differences perceptible to human observers <ref:2607. p. 0
Where the paper says it

Engy Ehab 1,2,4 , Pablo Hernandez-C ´ amara ´ 3 , Nahla Belal 4 , Jes ´us Malo 3 , Javier Vazquez-Corral, 1,2 , Alexandra Gomez-Villa 1,2 1 Computer Vision Center, Spain 2 Universitat Autonoma de Barcelona, Spain ` 3 Image Processing Lab, Universitat de Valencia, Spain ` 4 Arab Academy for Science, Technology, and Maritime Transport, Egypt Abstract Understanding and characterizing human color perception is a longstanding research goal. One of the most traditional approaches is looking for the human color discrimination thresholds, the minimum chromatic differences perceptible to human observers.

16540#pg2>. In this work, we present a large-scale exploratory study probing the chromatic sensitivity of more than 50 pretrained vision encoders, including convolutional networks and vision transformers, against human discrimination thresholds <ref:2607. p. 0
Where the paper says it

Despite the widespread use of deep vision encoders, few studies have investigated whether their internal representations exhibit human-like discrimination thresholds. In this work, we present a large-scale exploratory study probing the chromatic sensitivity of more than 50 pretrained vision encoders, including convolutional networks and vision transformers, against human discrimination thresholds.

16540#pg2>. Our analysis reveals generally weak alignment between model representations and human perceptual thresholds across all model families, with the best mIoU .

Pruning redundant reasoning tokens

The study systematically evaluates the chromatic representation geometry of deep vision encoders against human discrimination behavior <ref:2607.16540#pg2>. This framework allows us to quantify the extent to which modern vision encoders align with human perceptual chromatic thresholds <ref:2607. p. 0

Where the paper says it

Using controlled chromatic stimuli and similarity fields derived from latent feature embeddings, we compare model responses against human discrimination thresholds defined in CIEDE2000 space. This framework allows us to quantify the extent to which modern vision encoders align with human perceptual chromatic thresholds.

16540#pg2>. The evaluation involves comparing model-derived chromatic discrimination thresholds with human discrimination ellipses through a region-overlap metric (mIoU) <ref:2607. p. 2
Where the paper says it

By comparing the structure of these similarity fields with human discrimination thresholds, we can evaluate the extent to which modern vision encoders exhibit human-aligned chromatic thresholds. Evaluation Metrics To quantify alignment between model-derived similarity fields and human discrimination thresholds, we compute a region-overlap metric based on intersection over union (IoU, Eq.

16540#pg2>.

How the training signal is built

The researchers evaluate a large and heterogeneous collection of 50 pretrained vision encoders spanning convolutional neural networks, vision transformers, and hybrid architectures <ref:2607.16540#pg2>. The benchmark includes multiple variants within each model family, with variability arising from differences in scale, training objectives, and optimization strategies <ref:2607. p. 1

Where the paper says it

Methodology Vision Encoders We evaluate a large and heterogeneous collection of 50 pretrained vision encoders spanning a broad spectrum of contemporary visual representation learning paradigms. Rather than relying on a single representative per architecture, the benchmark includes multiple variants within each model family, with variability arising from differences in scale, training objectives, and optimization strategies.

16540#pg2>.

Where the gains come from

The results indicate that self-supervised encoders generally achieve the strongest alignment scores <ref:2607.16540#pg2>. Self-supervised objectives preserve low-level chromatic relationships more effectively, suggesting that objectives preserving local image structure and visual consistency may better retain more perceptually human-like chromatic thresholds <ref:2607. p. 3

Where the paper says it

Nevertheless, the results reveal trends across training paradigms. Self-supervised encoders generally achieve the strongest alignment scores, suggesting that objectives preserving local image structure and visual consistency may better retain more perceptually human-like chromatic thresholds.

16540#pg2>. Languagesupervised models show the most polarized behavior, occupying both the top and bottom of the ranking <ref:2607. p. 0
Where the paper says it

Our analysis reveals generally weak alignment between model representations and human perceptual thresholds across all model families, with the best mIoU < 0. Moreover, we find that self-supervised encoders consistently outperform supervised ones, while languagesupervised models show the most polarized behavior, occupying both the top and bottom of the ranking.

16540#pg2>.

Dependence on Hue Region and Chroma Level

The alignment is highly dependent on both chromatic location and saturation regime <ref:2607.16540#pg2>. Certain hue regions, particularly blues, consistently produce lower IoU scores across nearly all encoders <ref:2607. p. 3

Where the paper says it

The results reveal that alignment is highly dependent on both chromatic location and saturation regime. Certain hue regions (particularly blues, which are also known to exhibit strong perceptual non-uniformities in classical psychophysical datasets) consistently produce lower IoU scores across nearly all encoders, indicating that these areas of color space are systematically more difficult for learned representations to organize in a psychophysically consistent manner.

16540#pg2>. This behavior is especially pronounced at high chroma levels, where the human discrimination ellipses are larger and their anisotropic structure more complex <ref:2607. p. 3
Where the paper says it

Certain hue regions (particularly blues, which are also known to exhibit strong perceptual non-uniformities in classical psychophysical datasets) consistently produce lower IoU scores across nearly all encoders, indicating that these areas of color space are systematically more difficult for learned representations to organize in a psychophysically consistent manner. This behavior is especially pronounced at high chroma levels, where the human discrimination ellipses are larger and their anisotropic structure more complex.

16540#pg2>.

Conclusions

The results show that while deep vision encoders consistently develop structured and anisotropic chromatic similarity geometries, these representations only partially align with human perceptual models <ref:2607.16540#pg2>. Alignment varies systematically across chromaticity regions and saturation levels, and depends strongly on training paradigm, with self-supervised and supervised models outperforming language-supervised and generative approaches <ref:2607. p. 4

Where the paper says it

Our results show that while deep vision encoders consistently develop structured and anisotropic chromatic similarity geometries, these representations only partially align with human perceptual models. Alignment varies systematically across chromaticity regions and saturation levels, and depends strongly on training paradigm, with self-supervised and supervised models outperforming language-supervised and generative approaches.

16540#pg2>. However, even the best-performing models remain far from human-level chromatic discrimination <ref:2607. p. 4
Where the paper says it

Alignment varies systematically across chromaticity regions and saturation levels, and depends strongly on training paradigm, with self-supervised and supervised models outperforming language-supervised and generative approaches. However, even the best-performing models remain far from human-level chromatic discrimination, suggesting that humanlike chromatic discrimination does not naturally emerge from current large-scale visual training objectives.

16540#pg2>.

Limitations and future work

This study has some limitations, first focusing on static, synthetic stimuli designed to isolate chromatic effects under controlled conditions <ref:2607.16540#pg2>. Extending the evaluation to include structured stimuli, such as textured surfaces and object-based renderings, would help determine whether chromatic alignment improves under more ecologically valid conditions <ref:2607. p. 5

Where the paper says it

this enables precise comparison with MacAdam human discrimination thresholds, it necessarily abstracts away from richer visual contexts present in natural scenes. Extending the evaluation to include structured stimuli, such as textured surfaces and object-based renderings, would help determine whether chromatic alignment improves under more ecologically valid conditions.

16540#pg2>. A promising direction is to quantify whether models preserve the directional structure of MacAdam-like ellipses across different spatial contexts, including full objects rather than flat color patches <ref:2607. p. 5
Where the paper says it

Extending the evaluation to include structured stimuli, such as textured surfaces and object-based renderings, would help determine whether chromatic alignment improves under more ecologically valid conditions. In particular, a promising direction is to quantify whether models preserve the directional structure of MacAdam-like ellipses across different spatial contexts, including full objects rather than flat color patches.

16540#pg2>. Additionally, the evaluation is based on cosine distance in frozen feature spaces, which may not fully capture nonlinear decoding processes in downstream tasks <ref:2607. p. 5
Where the paper says it

Such an analysis would help disentangle whether current models encode chromatic sensitivity primarily at the pixel level or whether these properties persist in higher-level representations. Additionally, our evaluation is based on cosine distance in frozen feature spaces, which may not fully capture nonlinear decoding processes in downstream tasks.

16540#pg2>.

Acknowledgments

The work was partially funded by Grants PID2023-152133NB-I00, PID2024-162555OB-I00; by MICIU/AEI/10.13039/501100011033, ERDF/EU and the FEDER; by Spanish MIU under FPU21/02256; by the grant Catedra ENIA UAB-Cru¨ılla (TSI-100929-2023- 2) from the Ministry of Economic Affairs and Digital Transition of Spain; by the grant BBVA Foundations of Science program: Maths, Stats, Comp. p. 5

Where the paper says it

Acknowledgments The work was partially funded by Grants PID2023- 152133NB-I00, PID2024-162555OB-I00; by MICIU/AEI/10. 13039/501100011033, ERDF/EU and the FEDER; by Spanish MIU under FPU21/02256; by the grant Catedra ` ENIA UAB-Cru¨ılla (TSI-100929-2023- 2) from the Ministry of Economic Affairs and Digital Transition of Spain; by the grant BBVA Foundations of Science program: Maths, Stats, Comp.

Sci. and AI (VIS4NN) JVC also acknowledges the 2025 Leonardo Grant for Scientific Research and Cultural Creation from the BBVA Foundation <ref:2607. p. 5
Where the paper says it

and AI (VIS4NN). JVC also acknowledges the 2025 Leonardo Grant for Scientific Research and Cultural Creation from the BBVA Foundation.

16540#pg2>. The BBVA Foundation accepts no responsibility for the opinions, statements and contents included in the project and/or the results thereof, which are entirely the responsibility of the authors <ref:2607. p. 5
Where the paper says it

JVC also acknowledges the 2025 Leonardo Grant for Scientific Research and Cultural Creation from the BBVA Foundation. The BBVA Foundation accepts no responsibility for the opinions, statements and contents included in the project and/or the results thereof, which are entirely the responsibility of the authors.

16540#pg2>.

References

[1] David L MacAdam. Visual sensitivities to color differences in daylight <ref:2607. p. 5

Where the paper says it

References [1] David L MacAdam. Visual sensitivities to color differences in daylight.

16540#pg2>

[2] Ming Ronnier Luo and Bob Rigg. Chromaticity-discrimination ellipses for surface colours <ref:2607. p. 5

Where the paper says it

[2] Ming Ronnier Luo and Bob Rigg. Chromaticity-discrimination ellipses for surface colours.

16540#pg2>

[3] Roy S. Berns, David H. p. 5

Where the paper says it

Color Research & Application, 11(1):25–42, 1986. Berns, David H.

Alman, Lisa Reniff, G. p. 5
Where the paper says it

Berns, David H. Alman, Lisa Reniff, G.

D. Snyder, and M. p. 5
Where the paper says it

Alman, Lisa Reniff, G. Snyder, and M.

R. Balonon-Rosen <ref:2607. p. 5
Where the paper says it

Snyder, and M. Balonon-Rosen.

16540#pg2>

[4] Commission Internationale de l’Eclairage Colorimetry <ref:2607.16540#pg2>

[5] Roy S. Berns Billmeyer and Saltzman’s Principles of Color Technology <ref:2607. p. 5

Where the paper says it

15, Central Bureau of the CIE, Vienna, Austria, 1976. Billmeyer and Saltzman’s Principles of Color Technology.

16540#pg2>

[6] F. J. J. Clarke, R. McDonald, and B. p. 5

Where the paper says it

John Wiley & Sons, Hoboken, NJ, 4th edition, 2019. McDonald, and B.

Rigg Modification to the JPC79 colour-difference formula <ref:2607. p. 5
Where the paper says it

McDonald, and B. Modification to the JPC79 colour-difference formula.

16540#pg2>

[7] Commission Internationale de l’Eclairage Industrial colour-difference evaluation <ref:2607.16540#pg2>

[8] M Ronnier Luo, Guihua Cui, and Bryan Rigg The development of the cie 2000 colour-difference formula Ciede2000 <ref:2607.16540#pg2>

[9] Alexandra Gomez-Villa Kai Wang C. Alejandro Parraga Bartlomiej Twardowski Jesus Malo Javier Vazquez-Corral and Joost van den Weijer The art of deception Color visual illusions and diffusion models <ref:2607. p. 5

Where the paper says it

Alejandro Parraga, Bartlomiej Twardowski, Jesus Malo, Javier Vazquez-Corral, and Joost van den Weijer. The art of deception: Color visual illusions and diffusion models.

16540#pg2>

[10] Alex Gomez-Villa Adrian Mart´ın Javier Vazquez-Corral Marcelo Bertalm´ıo and Jesus Malo Color illusions also deceive cnns for low-level vision tasks Analysis and implications <ref:2607.16540#pg2>

[11] Jaykishan Patel Alban Flachot Javier Vazquez-Corral David H. Brainard Thomas S. p. 5

Where the paper says it

[11] Jaykishan Patel, Alban Flachot, Javier Vazquez-Corral, David H. Brainard, Thomas S.

A. Wallis Marcus A. p. 5
Where the paper says it

Brainard, Thomas S. Wallis, Marcus A.

Brubaker and Richard F. p. 5
Where the paper says it

Wallis, Marcus A. Brubaker, and Richard F.

Murray A deep convolutional neural network trained for lightness constancy is susceptible to lightness illusions <ref:2607. p. 5
Where the paper says it

Brubaker, and Richard F. A deep convolutional neural network trained for lightness constancy is susceptible to lightness illusions.

16540#pg2>

[12] Guillem Arias Ariadna Sola Mart ´ı Armengod, and Maria Vanrell Color encoding in latent space of stable diffusion models <ref:2607.16540#pg2>

[13] Yancheng Cai Fei Yin Dounia Hammou, and Rafal Mantiuk Do computer vision foundation models learn the low-level characteristics of the human visual system In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition CVPR 2025 <ref:2607.16540#pg2>

[14] Nipun Ravindu Wickramanayaka and Masafumi Oizumi Systematic comparison of color representations between humans and deep neural networks towards predicting human color perception in a vast color space bioRxiv pages 2025–12, 2025 <ref:2607.16540#pg2>

[15] Alec Radford Jong Wook Kim Chris Hallacy Aditya Ramesh, et al. Learning transferable visual models from natural language supervision arXiv preprint arXiv:2103. p. 5

Where the paper says it

Learning transferable visual models from natural language supervision. arXiv preprint arXiv:2103.

00020, 2021 <ref:2607. p. 5
Where the paper says it

arXiv preprint arXiv:2103. 00020, 2021.

16540#pg2>

[16] Gabriel Ilharco et al. Openclip 2021 GitHub repository <ref:2607. p. 5

Where the paper says it

Openclip, 2021. GitHub repository.

16540#pg2>

[17] Maxime Oquab Timothee Darcet Theo Moutakanni, et al. Dinov2 Learning robust visual features without supervision arXiv preprint arXiv:2304. p. 5

Where the paper says it

Dinov2: ´ Learning robust visual features without supervision. arXiv preprint arXiv:2304.

07193, 2023 <ref:2607. p. 5
Where the paper says it

arXiv preprint arXiv:2304. 07193, 2023.

16540#pg2>

[18] Stability AI Stable diffusion 3 https://stability.ai, 2024 <ref:2607. p. 0

Where the paper says it

They found that some models, particularly DINOv2 and OpenCLIP, exhibit partial alignment with human vision, although substantial differences remain Closer to our work, Wickramanayaka and Oizumi [14] systematically compared the internal color representations of 16 DNNs against human color similarity judgments of 93 colors, finding that only CLIP models achieved fine-grained structural congruence with human perceptual data. Crucially, however, their analysis targets the global representational geometry of color space — whether arXiv:2607.

16540#pg2>

[19] X. Zhang et al. p. 5

Where the paper says it

https://stability. Zhang et al.

Pe-clip Perception-enhanced clip for image-text representation learning arXiv preprint, 2023 <ref:2607. p. 5
Where the paper says it

Pe-clip: Perception-enhanced clip for image-text representation learning. arXiv preprint, 2023.

16540#pg2>

[20] Karen Simonyan and Andrew Zisserman Very deep convolutional networks for large-scale image recognition arXiv preprint arXiv:1409.1556, 2014 <ref:2607. p. 5

Where the paper says it

arXiv preprint arXiv:1409. 1556, 2014.

16540#pg2>

[21] Gaurav Sharma Wencheng Wu and Edul N Dalal The ciede2000 color-difference formula Implementation notes supplementary test data and mathematical observations Color Research & Application, 30(1):21–30, 2005 <ref:2607.16540#pg2>

[22] IEC A standard default color space for the internet - srgb, 1996 IEC 61966-2-1 <ref:2607.16540#pg2>

[23] Alexey Dosovitskiy et al. An image is worth 16x16 words Transformers for image recognition at scale arXiv preprint arXiv:2010. p. 5

Where the paper says it

An image is worth 16x16 words: Transformers for image recognition at scale. arXiv preprint arXiv:2010.

11929, 2020 <ref:2607. p. 5
Where the paper says it

arXiv preprint arXiv:2010. 11929, 2020.

Green — the paper says it, and the page is printed. Red — no passage carries it, so treat that sentence as unverified. The same check runs on your own text: check a summary against its source, $5.

Improvements for AI systems

  1. Improve color fidelity in digital art preservation and fine-grained product inspection by using self-supervised models, such as DINOv2 or Perception Encoder, for latent space representations because "Self-supervised encoders generally achieve the strongest alignment scores, suggesting that objectives preserving local image structure and visual consistency may better retain more perceptually human-like chromatic thresholds."

  2. Develop color analysis tools that are robust across different saturation regimes by prioritizing models trained with self-supervised objectives over purely supervised ones, as the results show the superiority of self-supervised models over purely supervised shows a clear dependence and the importance of the optimization goal.

  3. Enhance model generalization in color tasks by utilizing language-supervised models like AIMv2 when seeking top performance, as these models show the most polarized behavior, occupying both the top and bottom of the ranking, suggesting their training objectives capture broader organizational principles.

  4. Implement a systematic method for diagnosing latent space failures by using cosine distance similarity fields to identify where the precise geometric organization of the learned fields remains substantially misaligned with human references, allowing developers to pinpoint specific color-space regions where models fail to reproduce perceptual non-uniformities.

  5. Design color processing pipelines that account for chromatic location dependence by dynamically adjusting encoding based on hue and chroma levels, as the study found that alignment is highly dependent on both chromatic location and saturation regime and noted that certain hue regions (particularly blues, which are also known to exhibit strong perceptual non-uniformities in classical psychophysical datasets) consistently produce lower IoU scores.

Abstract

Understanding and characterizing human color perception is a longstanding research goal. One of the most traditional approaches is looking for the human color discrimination thresholds, the minimum chromatic differences perceptible to human observers. In recent years, deep neural networks have become the standard networks for computer vision tasks. In particular, deep vision encoders, foundation models trained on large-scale visual data, map images into latent feature representations. Despite the widespread use of deep vision encoders, few studies have investigated whether their internal representations exhibit human-like discrimination thresholds. In this work, we present a large-scale exploratory study probing the chromatic sensitivity of more than 50 pretrained vision encoders, including convolutional networks and vision transformers, against human discrimination thresholds. Using controlled chromatic stimuli at multiple chroma levels, we compare model-derived chromatic discrimination thresholds with human discrimination ellipses through a region-overlap metric (mIoU). Our analysis reveals generally weak alignment between model representations and human perceptual thresholds across all model families, with the best mIoU < 0.25. Moreover, we find that self-supervised encoders consistently outperform supervised ones, while language-supervised models show the most polarized behavior, occupying both the top and bottom of the ranking. These findings suggest that human-like chromatic sensitivity does not emerge naturally from current large-scale visual training objectives for any of the analyzed architectures.

Sources

Related papers