Related Experiment Videos
A model for contrast discrimination with incremental and decremental test patches
Vision Research
|January 1, 1991
Summary
This study proposes a new contrast metric, G, as a more physiologically plausible alternative to Whittle's metric W for describing visual contrast discrimination. Metric G offers a better fit to experimental data and implications for retinal gain mechanisms.
Area of Science:
- Visual perception
- Psychophysics
- Computational neuroscience
Background:
- Whittle's contrast metric (W) unifies contrast discrimination thresholds but lacks functional plausibility.
- Metric W implies different adaptational levels for luminance increments and decrements.
Purpose of the Study:
- To evaluate the physiological plausibility of Whittle's contrast metric (W).
- To introduce and validate a new, more plausible contrast metric (G).
- To explore implications for retinal gain mechanisms.
Main Methods:
- Analysis of Whittle's contrast discrimination data.
- Comparison of mathematical fits for metrics W and G.
- Incorporation of metric G into a transfer function (RG = kG1-n).
- Inclusion of a non-linearity parameter for improved fit.
Main Results:
- Metric W, while mathematically descriptive, is functionally implausible.
- Metric G (ln(L/Lb)) provides a physiologically plausible alternative.
- Metric G, within a transfer function with n=0.69, fits Whittle's data as well as W.
- Including a low-luminance non-linearity parameter further improves the fit of metric G.
Conclusions:
- Metric G is a more physiologically plausible descriptor of contrast discrimination than metric W.
- The proposed model with metric G has significant implications for understanding retinal gain control.
- Further research into non-linear gain mechanisms in the retina is warranted.