castleCSF - A contrast sensitivity function of color, area, spatiotemporal frequency, luminance and eccentricity

Area of Science:

• Vision science
• Perceptual modeling
• Engineering psychology

Background:

• The contrast sensitivity function (CSF) is crucial for understanding visual perception and has engineering applications.
• Existing CSFs lack comprehensive explanations for diverse stimulus parameters.
• A unified model is needed to cover spatial, temporal, luminance, and chromatic factors.

Purpose of the Study:

• To consolidate existing contrast sensitivity data from multiple studies.
• To develop a novel analytical model (castleCSF) for predicting contrast sensitivity.
• To provide a publicly available dataset and model for research and application.

Main Methods:

• Collected and integrated contrast sensitivity measurements from 18 diverse studies.
• Developed an analytical model incorporating spatial/temporal frequencies, color modulation, luminance, chromaticity, eccentricity, and area.
• Validated the model against the consolidated dataset.

Main Results:

• The castleCSF model accurately explains contrast sensitivity across a wide range of stimulus parameters.
• The model achieves a low error of 3.59 dB when fitting data from 18 studies.
• Consolidated data and model code are publicly accessible.

Conclusions:

• The castleCSF model offers a unified and accurate prediction of human contrast sensitivity.
• This model advances applications in visual engineering and perceptual design.
• Public data and code facilitate further research and development in visual perception.