Scholarlink

<p>Realism remains an unresolved question in vision science. This study investigates whether the physical gamut correlates with realism judgements. We conducted psychophysical experiments where observers judged the realism of natural scenes with target regions manipulated across the CIE 1931 color space. Results initially showed a moderate-to-strong correlation between judgements and a theoretical physical gamut derived from optimal colors. Further analysis revealed that the most detrimental points were in the saturated green region of the CIE 1931 <i>xy</i> chromaticity diagram; removing them yielded a very strong correlation. To explain this discrepancy, we modeled a real-world physical gamut based on USGS and ECOSTRESS spectral libraries, which compile thousands of spectra from natural and man-made materials. The analysis revealed that the detrimental green chromaticities might be non-existent in the real world. Since physical gamut theory posits that the visual system constructs internal references through empirical observation of the world, the absence of these colors in nature might be an explanation to the theoretical model’s failure. Ultimately, the real-world gamut exhibited an even stronger correlation with judgements, supporting our hypothesis while suggesting that the theoretical model may not be the optimal approximation of the actual physical gamut. These findings contribute to discussions on perceptual realism.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

The perception of realism is correlated with physical gamuts

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年8月5日

查看全文
  • Killian Duay,
  • Takehiro Nagai

摘要

Realism remains an unresolved question in vision science. This study investigates whether the physical gamut correlates with realism judgements. We conducted psychophysical experiments where observers judged the realism of natural scenes with target regions manipulated across the CIE 1931 color space. Results initially showed a moderate-to-strong correlation between judgements and a theoretical physical gamut derived from optimal colors. Further analysis revealed that the most detrimental points were in the saturated green region of the CIE 1931 xy chromaticity diagram; removing them yielded a very strong correlation. To explain this discrepancy, we modeled a real-world physical gamut based on USGS and ECOSTRESS spectral libraries, which compile thousands of spectra from natural and man-made materials. The analysis revealed that the detrimental green chromaticities might be non-existent in the real world. Since physical gamut theory posits that the visual system constructs internal references through empirical observation of the world, the absence of these colors in nature might be an explanation to the theoretical model’s failure. Ultimately, the real-world gamut exhibited an even stronger correlation with judgements, supporting our hypothesis while suggesting that the theoretical model may not be the optimal approximation of the actual physical gamut. These findings contribute to discussions on perceptual realism.