LEARN THE LATEST ABOUT COLOR

 

Mingxue Wang is the director of media technology and the standards lab in Huawei’s Central Media Technology Institute. He obtained his MS degree from Zhejiang University. After joining Huawei, he has worked in different departments and been deeply involved in R&D work on media technologies and standardizations.

It is impossible to recover the actual reflectance that induces a given colour response: as many spectra - called metamers - will integrate to the same response values. For some applications it suffices to recover a good single metamer (satisfying a criterion such that it is the smoothest amongst all metamers). However, when the same surface is viewed under different lights - generating different RGBs - the corresponding reflectances recovered by Smoothest Reflectance estimation (SR) are not all the same. Indeed, there can be a large spectral variation. Recent work has demonstrated that more stable - illuminant insensitive - metamers can be produced by Colour Corrected Smoothest Reflectance estimation (CCSR): where camera RGBs are colour corrected to a canonical reflectance light with respect to which metamers are recovered. In this paper, we examine the relationship between the spectral sensitivities of the camera and both SR and CCSR metamer recovery. Empirically, the variation in recovered metamers for the worst camera for the SR method is found to be 2.5 times larger than the best camera using CCSR. We argue that the stability of metamer recovery in general (for either SR or CCSR) is linked to the extent that accurate colour correction is possible.

We investigated how to optimally adjust brand logo colors for dark mode displays without compromising their original color identity. Through analyzing manual color adjustments of logos placed on black backgrounds (N=31), we figured out that designers tend to light up the dark-colored logo, while pale down the colors of bright logo. A tendency similar to that of their adjustments converging in a specific direction was obtained, so we derived a surface model that represents the direction in which original colors tend to shift. This model incorporates the L*, C*, and h* attributes to guide the computational adjustment process. This study offers a structured method for adapting logo colors to contemporary display contexts, effectively linking algorithmic solutions with design intuition.

Automatic white balance (AWB) plays a crucial role in digital imaging, with modern learning-based methods achieving better performance. These methods, however, require extensive training data captured by a specific sensor, which cannot be directly deployed other sensors due to the different spectral sensitivity functions. This paper presents a novel cross-sensor adaptation method based on 3×3 color transformation matrices. By leveraging least-squares optimization and a Mahalanobis distance strategy, our approach constructs sensor-specific mapping matrices using 24-patch ColorChecker data. The results derived using the NUS dataset demonstrate that the proposed method has much smaller angular errors without requiring additional data collection or complicated network tuning.

Accurate white balancing (WB) remains a critical challenge in image signal processing. Particularly white point estimation for scenes with limited or ambiguous color information can lead to color casts and degraded visual quality. Moreover, many every-day scenes contain multiple relevant illuminants, further complicating illuminant estimation. Estimating multiple white points in a scene exacerbates the challenge and traditional RGB-based WB algorithms struggle with the limited information available for localized regions of a scene. We introduce a novel approach leveraging compact multi-spectral camera technology to improve local WB performance. By capturing additional, narrow-band spectral information beyond the RGB channels, our method provides more accurate white point estimation. We present comparative results demonstrating the advantages of multi-spectral sensing over conventional approaches, highlighting its potential to enable more intelligent and adaptive imaging pipelines in mobile, automotive, and industrial applications. Our approach is based on a relatively simple neural network, trained on simulated multi-spectral measurements. We developed a data collection protocol to establish a medium-sized validation dataset for which we report white point angles and color accuracy (DE2000) values. Our study includes a performance benchmark against a state-of-the art deep learning-based algorithm for our new validation set and the publicly available Large Scale Multi-Illuminant (LSMI) dataset. From the results we observe that our network performs on par with the state-of-the-art algorithm on the LSMI dataset and outperforms the state-of-the-art algorithm on our validation dataset. Furthermore, the multi-spectral based network outperforms the RGB based network.

Shoji Tominaga received the Ph.D. in electrical engineering from Osaka University, Japan, in 1975. In 2006, he joined Chiba University, Japan, where he was a professor (2006-2013), dean (2011-2013), and a specially appointed researcher (2013-2018) at the Graduate School of Advanced Integration Science. He is an adjunct professor at NTNU, Norway, and a visiting researcher at Nagano University, Japan. His research interests include multispectral imaging and material appearance. He is a life fellow of IEEE, a fellow of IS&T, a life fellow of SPIE, and a fellow OSA.

Color-accurate digital imaging is critical for agricultural phenotyping, but the scientific literature predominantly assumes the availability of linear RAW sensor data. In some commercial workflows, only standard 8-bit sRGB JPEG images are available, which poses a significant challenge due to their non-linear encoding and information loss from in-camera processing. This paper presents a robust color correction pipeline designed specifically for such non-linear images captured in uncontrolled outdoor environments.

Our core contribution is a novel, per-patch adaptive weighting scheme for least-squares color correction. Instead of deriving a single global correction, our method generates a unique transformation for each patch on an in-frame ColorChecker. This is achieved through a leave-one-out approach where, for each target patch, a model is trained on the remaining 23 patches. Crucially, this training is guided by a weighting matrix customized for the target patch. This adaptive process allows a simple linear model to outperform more complex polynomials. Through systematic evaluation on an unseen test set, we demonstrate this method reduces the mean color error ΔE00 from 11.23 to 3.79, providing a practical and effective solution for real-world agricultural imaging.

 

Hyeon-Jeong Suk is professor and head of the Department of Industrial Design at KAIST. She also serves as director of the KAIST Art Museum and leads the university’s brand business unit. Suk received her BSc and MSc in Industrial Design from KAIST, and earned her PhD in Psychology from the University of Mannheim in Germany. Her research focuses on color psychology, human-centered lighting, and color cosmetics, bridging color science and design practice to create value-driven experiences.

Color charts are used in many color imaging contexts to sample a color space or the effects of a color transformation. A rich variety of approaches exist here for determining the set of colors contained in a color chart and the specific choices of colors and their cardinality both then impact the goodness of the color imaging processes that use them. A key example here are the color charts used for building ICC color profiles. Such charts are often variations on uniform samplings in either an RGB or a CMYK device color space. At the same time, the accuracy of an ICC profile is judged in terms of color difference (ΔEs), and a chart that is uniformly distributed in those terms would be advantageous. This paper presents an approach to explicitly tuning color charts for uniformity in arbitrary color spaces and for arbitrary color difference metrics, using computational optimization over a hexagonally close packed mesh. Results show a substantial improvement in the uniformity of ΔE differences between each chart color and its neighbors.

A controlled experimental setup used multichannel LED lighting to create HDR scenes. Ten observers performed visual matching tasks between real illuminated scenes and HDR display content across eight lighting conditions. J_za_zb_z, CIECAM16, and CAM16-UCS were evaluated, analysis using STRESS metrics showed CAM16-UCS achieved the best performance for both lightness and colorfulness predictions. Based on these findings, a tone mapping operator was developed utilizing CAM16-UCS color space with local adaptation and gamma adjustments derived from experimental data. The results demonstrate that CAM16-UCS provides superior color appearance prediction for HDR content and serves as an effective foundation for tone mapping applications.

Tone mapping algorithms are used to compress dynamic range, make image details more conspicuous and generally enhance the image for preference. Global tone mapping manipulates the brightnesses of pixels by applying a single function - or tone curve - to every pixel in the image. Tone curve generation algorithms often constrain the shape of their tone curves and it has been argued that tone curves should be simple, meaning they have one or zero inflexion points. In this work, we investigate whether tone curves should be simplified even further. We present our method which finds the zero inflexion tone curve - which we call a Very Simple (VS) curve - that best approximates a potentially complex tone curve. For the MIT-Adobe FiveK dataset, comprising 25,000 expert tone adjustments, we calculate the best VS approximations and find these curves produce visually similar images compared with more complex counterparts.

Despite decades of research and significant investment, digital imaging technology still struggles in certain real-world scenarios—such as capturing a photo at an evening gathering in a dimly lit restaurant—where either shadows appear completely dark or bright areas become oversaturated, losing vital details. Even with modern HDR (High Dynamic Range) features now standard in smartphones, these issues often persist. And yet, the human eye can perceive detail across both shadowed and brightly lit areas in such scenes with ease.

This talk explores the fundamental reason behind this discrepancy: the inherent limitations of digital sensors in handling high dynamic range light intensities. Professor Qiu will revisit the HDR problem from first principles, arguing that it remains one of the most critical—and often misunderstood—challenges in digital imaging. He will critique common misconceptions in recent literature, particularly those relying on deep learning-based “black-box” approaches to low-light enhancement. Through theoretical insights and practical demonstrations, the talk will make the case that HDR remains the last great frontier in the evolution of digital photograph.

Three main aspects of the appearance of metallic coatings are color, gloss and texture (including graininess and sparkle). All three aspects can now be objectively measured using instruments according to ASTM and other international standards. However, there is no reason why these three aspects would be independent from each other. Indeed, it is known for more than a century that variations in color influence the perception of gloss. Here, we present the first systematic investigation on the influence of gloss variations on the occurrence of sparkle and graininess.

We created a new, dedicated set of 462 solid and metallic paint samples, with systematic variations in color and gloss. We measured multi-angle reflectance properties for all samples. Our results confirm that color strongly influences measured gloss values, especially for metallic coatings. We also show that there is a substantial influence from glossiness to the measured graininess and sparkle properties. For example, by applying a matte clearcoat instead of a more common high-gloss clearcoat on top of a metallic basecoat, the measured graininess reduces by 0.6±0.3 units and the sparkle grade at 15° by 3.3±1.7 units. This is in line with visual assessments on matte car paints: matte metallic paints hardly show sparkle. A physical explanation for these trends is presented.

One central challenge in modeling material appearance perception is the creation of an explainable and navigable representation space. In this study, we address this by training a StyleGAN2-ADA deep generative model on a large-scale, physically-based rendered dataset containing translucent and glossy objects with varying intrinsic optical parameters. The resulting latent vectors were analyzed through dimensionality reduction, and their perceptual validity was assessed via psychophysical experiments. Furthermore, we evaluated the generalization capabilities of StyleGAN2-ADA on unseen materials. We also explore inverse mapping techniques from PCA-reduced latent vectors back to original optical parameters, highlighting both the potential and the limitations of generative models for explicit, parameter-based image synthesis. Our comprehensive analysis provides significant insights into the latent structure of gloss and translucency perception and advances the practical application of generative models for controlled material appearance generation.