Abstract

The ideal-observer performance, as measured by the area under the receiver’s operating characteristic curve, is computed for six examples of signal-detection tasks. Exact values for this quantity, as well as approximations based on the signal-to-noise ratio of the log likelihood and the likelihood-generating function, are found. The noise models considered are normal, exponential, Poisson, and two-sided exponential. The signal may affect the mean or the variance in each case. It is found that the approximation from the likelihood-generating function tracks well with the exact area, whereas the log-likelihood signal-to-noise approximation can fail badly. The signal-to-noise ratio of the likelihood ratio itself is also computed for each example to demonstrate that it is not a good measure of ideal-observer performance.

© 2000 Optical Society of America

Using Fisher information to approximate ideal-observer performance on detection tasks for lumpy-background images

Fangfang Shen and Eric Clarkson
J. Opt. Soc. Am. A 23(10) 2406-2414 (2006)

Objective assessment of image quality. III. ROC metrics, ideal observers, and likelihood-generating functions

Harrison H. Barrett, Craig K. Abbey, and Eric Clarkson
J. Opt. Soc. Am. A 15(6) 1520-1535 (1998)

Channelized-ideal observer using Laguerre-Gauss channels in detection tasks involving non-Gaussian distributed lumpy backgrounds and a Gaussian signal

Subok Park, Harrison H. Barrett, Eric Clarkson, Matthew A. Kupinski, and Kyle J. Myers
J. Opt. Soc. Am. A 24(12) B136-B150 (2007)

Asymptotic ideal observers and surrogate figures of merit for signal detection with list-mode data

Eric Clarkson
J. Opt. Soc. Am. A 29(10) 2204-2216 (2012)

Estimation receiver operating characteristic curve and ideal observers for combined detection/estimation tasks

Eric Clarkson
J. Opt. Soc. Am. A 24(12) B91-B98 (2007)