Too Vivid to Be Real? Benchmarking and Calibrating Generative Color Fidelity

• URL: http://arxiv.org/abs/2603.10990v1
• Date: Wed, 11 Mar 2026 17:18:12 GMT
• Title: Too Vivid to Be Real? Benchmarking and Calibrating Generative Color Fidelity
• Abstract summary: We present Color Fidelity dataset (CFD) and Color Fidelity Metric (CFM) for objective evaluation of color fidelity in realistic-style generations.<n> CFD contains over 1.3M real and synthetic images with ordered levels of color realism, while CFM employs a multimodal encoder to learn perceptual color fidelity.<n>We propose a training-free Color Fidelity Refinement (CFR) that adaptively modulates spatial-temporal guidance scale in generation, thereby enhancing color authenticity.
• Score: 55.22847424233036
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Recent advances in text-to-image (T2I) generation have greatly improved visual quality, yet producing images that appear visually authentic to real-world photography remains challenging. This is partly due to biases in existing evaluation paradigms: human ratings and preference-trained metrics often favor visually vivid images with exaggerated saturation and contrast, which make generations often too vivid to be real even when prompted for realistic-style images. To address this issue, we present Color Fidelity Dataset (CFD) and Color Fidelity Metric (CFM) for objective evaluation of color fidelity in realistic-style generations. CFD contains over 1.3M real and synthetic images with ordered levels of color realism, while CFM employs a multimodal encoder to learn perceptual color fidelity. In addition, we propose a training-free Color Fidelity Refinement (CFR) that adaptively modulates spatial-temporal guidance scale in generation, thereby enhancing color authenticity. Together, CFD supports CFM for assessment, whose learned attention further guides CFR to refine T2I fidelity, forming a progressive framework for assessing and improving color fidelity in realistic-style T2I generation. The dataset and code are available at https://github.com/ZhengyaoFang/CFM.

Related papers

• ColorFM: An Optimization-to-Learning Framework for Color Transfer via Flow Matching [38.73819453939708]
  We propose ColorFM, an optimization-to-learning framework for color transfer.<n> ColorFM connects online optimization to offline inference by reformulating color transfer as the transport of pixel distributions.<n>We show that ColorFM-L outperforms state-of-the-art methods in visual quality, structural fidelity, and semantic consistency.
  arXiv  Detail & Related papers  (2026-07-08T08:01:15Z)
• CanonCGT: Reference-Based Color Grading via Canonical Pivot Representation [16.555135705482485]
  CanonCGT is a style-neutral intermediate representation for stable color mapping.<n>A dual-phase training scheme, DP-CGT, combines supervised preset learning with self-supervised refinement on unpaired photographs.
  arXiv  Detail & Related papers  (2026-06-01T03:42:20Z)
• Bringing Multimodal Large Language Models to Infrared-Visible Image Fusion Quality Assessment [48.882448108261826]
  Infrared-Visible image fusion (IVIF) aims to integrate thermal information and detailed spatial structures into a single fused image to enhance perception.<n>Recent IVIF reward-modelling efforts learn from human ratings but use scalar regression on aggregated scores.<n>We introduce FuScore, which utilizes an MLLM to mimic human visual perception by producing continuous quality score.
  arXiv  Detail & Related papers  (2026-05-07T21:38:09Z)
• SemiNFT: Learning to Transfer Presets from Imitation to Appreciation via Hybrid-Sample Reinforcement Learning [10.278887297409284]
  We propose a Diffusion Transformer-based retouching framework that mirrors the trajectory of human artistic training.<n>SemiNFT is first taught with paired triplets to acquire basic structural preservation and color mapping skills.<n>During the reinforcement learning stage, we design a hybrid online-offline reward mechanism that anchors aesthetic exploration with structural review.
  arXiv  Detail & Related papers  (2026-02-09T12:20:33Z)
• Wavelet-Domain Masked Image Modeling for Color-Consistent HDR Video Reconstruction [69.35623794013152]
  High Dynamic Range (LDR) video reconstruction aims to recover fine brightness, color, and details from LDR videos.<n>Existing methods often suffer from color inaccuracies and temporal inconsistencies.<n>We propose WMNet, a novel HDR video reconstruction network that leverages Wavelet domain Masked Image Modeling.
  arXiv  Detail & Related papers  (2026-02-07T06:19:23Z)
• GenColorBench: A Color Evaluation Benchmark for Text-to-Image Generation Models [61.786094845872576]
  We propose GenColorBench, the first comprehensive benchmark for text-to-image color generation.<n>It is grounded in color systems like I SCC-NBS and CSS3/X11, including numerical colors which are absent elsewhere.<n>With 44K color-focused prompts covering 400+ colors, it reveals models' true capabilities via perceptual and automated assessments.
  arXiv  Detail & Related papers  (2025-10-23T14:12:55Z)
• Structure-preserving Feature Alignment for Old Photo Colorization [19.011270929834495]
  We propose a novel CNN-based algorithm called SFAC, i.e., Structure-preserving Feature Alignment Colorizer.<n>SFAC is trained on only two images for old photo colorization, eliminating the reliance on big data.<n>Our primary objective is to establish semantic correspondence between the two images, ensuring that semantically related objects have similar colors.
  arXiv  Detail & Related papers  (2025-08-18T02:10:40Z)
• Leveraging Semantic Attribute Binding for Free-Lunch Color Control in Diffusion Models [53.73253164099701]
  We introduce ColorWave, a training-free approach that achieves exact RGB-level color control in diffusion models without fine-tuning.<n>We demonstrate that ColorWave establishes a new paradigm for structured, color-consistent diffusion-based image synthesis.
  arXiv  Detail & Related papers  (2025-03-12T21:49:52Z)
• Enhance Image Classification via Inter-Class Image Mixup with Diffusion Model [80.61157097223058]
  A prevalent strategy to bolster image classification performance is through augmenting the training set with synthetic images generated by T2I models. In this study, we scrutinize the shortcomings of both current generative and conventional data augmentation techniques. We introduce an innovative inter-class data augmentation method known as Diff-Mix, which enriches the dataset by performing image translations between classes.
  arXiv  Detail & Related papers  (2024-03-28T17:23:45Z)
• DiffBIR: Towards Blind Image Restoration with Generative Diffusion Prior [70.46245698746874]
  We present DiffBIR, a general restoration pipeline that could handle different blind image restoration tasks. DiffBIR decouples blind image restoration problem into two stages: 1) degradation removal: removing image-independent content; 2) information regeneration: generating the lost image content. In the first stage, we use restoration modules to remove degradations and obtain high-fidelity restored results. For the second stage, we propose IRControlNet that leverages the generative ability of latent diffusion models to generate realistic details.
  arXiv  Detail & Related papers  (2023-08-29T07:11:52Z)
• Improved Diffusion-based Image Colorization via Piggybacked Models [19.807766482434563]
  We introduce a colorization model piggybacking on the existing powerful T2I diffusion model. A diffusion guider is designed to incorporate the pre-trained weights of the latent diffusion model. A lightness-aware VQVAE will then generate the colorized result with pixel-perfect alignment to the given grayscale image.
  arXiv  Detail & Related papers  (2023-04-21T16:23:24Z)
• DDColor: Towards Photo-Realistic Image Colorization via Dual Decoders [19.560271615736212]
  DDColor is an end-to-end method with dual decoders for image colorization. Our approach includes a pixel decoder and a query-based color decoder. Our two decoders work together to establish correlations between color and multi-scale semantic representations.
  arXiv  Detail & Related papers  (2022-12-22T11:17:57Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.