Color brings relief to human vision

Area of Science:

• Visual Perception
• Computational Neuroscience
• Color Vision

Background:

• Natural scenes exhibit distinct color-luminance relationships: surface variations (flowers, paint) align color and luminance, while illumination variations (shadows) create luminance-only changes.
• The human visual system's processing of visual information is influenced by these natural statistical regularities.

Purpose of the Study:

• To investigate whether the human visual system utilizes learned color-luminance relationships for three-dimensional (3D) shape perception.
• To determine if the alignment between chromatic and luminance variations affects the perception of shape-from-shading.

Main Methods:

• Psychophysical experiments using 'plaid' stimuli composed of superimposed chromatic and luminance gratings.
• Varying the orientation alignment between the chromatic and luminance components of the plaid stimulus.
• Measuring the subjective perception of 3D shape-from-shading.

Main Results:

• Shape-from-shading perception was reliably triggered when the chromatic and luminance gratings were presented at different orientations.
• Conversely, shape-from-shading perception was suppressed when the chromatic and luminance gratings were aligned.
• This indicates a sensitivity of the visual system to the natural correlation between color and luminance.

Conclusions:

• The human visual system incorporates knowledge of natural color-luminance statistics to infer 3D shape from images.
• Color vision plays a crucial role in determining perceived 3D structure by leveraging these learned relationships.
• This finding offers new insights into the functional role of color in visual perception beyond simple object recognition.