OSA's Digital Library

Journal of the Optical Society of America A

Journal of the Optical Society of America A


  • Vol. 16, Iss. 3 — Mar. 1, 1999
  • pp: 764–778

Characterizing human perceptual inefficiencies with equivalent internal noise

Zhong-Lin Lu and Barbara Anne Dosher  »View Author Affiliations

JOSA A, Vol. 16, Issue 3, pp. 764-778 (1999)

View Full Text Article

Enhanced HTML    Acrobat PDF (1254 KB)

Browse Journals / Lookup Meetings

Browse by Journal and Year


Lookup Conference Papers

Close Browse Journals / Lookup Meetings

Article Tools



A widely used method for characterizing and comparing inefficiencies in perceptual processes is the method of equivalent internal noise—the amount of random internal noise necessary to produce the degree of inefficiency exhibited by the perceptual system in processing [J. Opt. Soc. Am. 46, 634 (1956)]. One normally estimates the amount of equivalent internal noise by systematically increasing the amount of external noise added to the signal stimulus and observing how threshold—signal stimulus energy required for an observer to maintain a given performance level—depends on the amount of external noise. In a variety of perceptual tasks, a simple noisy linear amplifier model [PelliD., Ph.D. dissertation (University of Cambridge, Cambridge, UK 1981)] has been utilized to estimate the equivalent internal noise Ninternal by fitting of the relation between threshold contrast cτ and external noise Next at a single (d) performance level: cτ2=(d/β)2(Next2+N internal2). This model makes a strong prediction: Independent of observer and external noise contrast, the ratio between two thresholds at each external noise level is equal to the ratio of the two corresponding d values. To our knowledge, this potential test for the internal consistency of the model had never been examined previously. In this study we estimated threshold ratios between multiple performance levels at various external noise contrasts in two different experiments: Gabor orientation identification, and Gabor detection. We found that, in both identification and detection, the observed threshold ratios between different performance levels departed substantially from the d ratio predicted by the simple noisy linear amplifier model. An elaborated perceptual template model [Vision Res. 38, 1183 (1998)] with nonlinear transducer functions and multiplicative noise in addition to the additive noise in the simple linear amplifier model leads to a substantially better description of the data and suggests a reinterpretation of earlier results that relied on the simple noisy linear amplifier model. The relationship of our model and method to other recent parallel and independent developments [J. Opt. Soc. Am. A 14, 2406 (1997)] is discussed.

© 1999 Optical Society of America

OCIS Codes
(330.1800) Vision, color, and visual optics : Vision - contrast sensitivity
(330.1880) Vision, color, and visual optics : Detection
(330.4060) Vision, color, and visual optics : Vision modeling
(330.5510) Vision, color, and visual optics : Psychophysics
(330.6100) Vision, color, and visual optics : Spatial discrimination
(330.7310) Vision, color, and visual optics : Vision

Original Manuscript: July 6, 1998
Revised Manuscript: November 12, 1998
Manuscript Accepted: December 1, 1998
Published: March 1, 1999

Zhong-Lin Lu and Barbara Anne Dosher, "Characterizing human perceptual inefficiencies with equivalent internal noise," J. Opt. Soc. Am. A 16, 764-778 (1999)

Sort:  Author  |  Year  |  Journal  |  Reset  


  1. H. B. Barlow, “Retinal noise and absolute threshold,” J. Opt. Soc. Am. 46, 634–639 (1956). [CrossRef] [PubMed]
  2. N. S. Nagaraja, “Effect of luminance noise on contrast thresholds,” J. Opt. Soc. Am. 54, 950–955 (1964). [CrossRef]
  3. D. G. Pelli, “Effects of visual noise,” Ph.D. dissertation (University of Cambridge, Cambridge, UK, 1981).
  4. A. E. Burgess, R. F. Wagner, R. J. Jennings, H. B. Barlow, “Efficiency of human visual signal discrimination,” Science 214, 93–94 (1981). [CrossRef] [PubMed]
  5. A. J. Ahumada, A. B. Watson, “Equivalent-noise model for contrast detection and discrimination,” J. Opt. Soc. Am. A 2, 1133–1139 (1985). [CrossRef] [PubMed]
  6. D. G. Pelli, “The quantum efficiency of vision,” in Vision: Coding and Efficiency, C. Blakemore, ed. (Cambridge U. Press, Cambridge, UK, 1990), pp. 3–24.
  7. D. H. Parish, G. Sperling, “Object spatial frequencies, retinal spatial frequencies, noise, and the efficiency of letter discrimination,” Vision Res. 31, 1399–1415 (1991). [CrossRef] [PubMed]
  8. G. Sperling, “Three stages and two systems of visual processing,” Spatial Vision 4, 183–207 (1989). [CrossRef] [PubMed]
  9. D. O. North, “The absolute sensitivity of radio receivers,” RCA Rev. 6, 332–344 (1942).
  10. H. T. Friis, “Noise figures of radio receivers,” Proc. IRE 32, 419–422 (1944). [CrossRef]
  11. W. W. Mumford, E. H. Schelbe, Noise Performance Factors in Communication Systems (Horizon House-Microwave, Dedham, Mass., 1968).
  12. H. Fletcher, “Auditory patterns,” Rev. Mod. Phys. 12, 47–65 (1940). [CrossRef]
  13. H. B. Barlow, “Incremental thresholds at low intensities considered as signal/noise discrimination,” J. Physiol. (London) 136, 469–488 (1957).
  14. J. A. Swets, D. M. Green, W. P. Tanner, “On the width of critical bands,” J. Acoust. Soc. Am. 34, 108–113 (1962). [CrossRef]
  15. U. Greis, R. Rohler, “Untersuchung der subjektiven Detailerkennbarkeit mit Hilfe der Ortsfrequenzfilterung,” Opt. Acta 17, 515–526 (1970). [CrossRef]
  16. H. Pollehn, H. Roehrig, “Effect of noise on the MTF of the visual channel,” J. Opt. Soc. Am. 60, 842–848 (1970). [CrossRef] [PubMed]
  17. B. E. Carter, G. B. Henning, “The detection of gratings in narrow-band visual noise,” J. Physiol. (London) 219, 355–365 (1971).
  18. C. F. Stromeyer, B. Julesz, “Spatial frequency masking in vision: critical bands and spread of masking,” J. Opt. Soc. Am. 64, 1221–1232 (1972). [CrossRef]
  19. L. D. Harmon, B. Julesz, “Masking in visual recognition: effects of two-dimensional filtered noise,” Science 180, 1194–1197 (1973). [CrossRef] [PubMed]
  20. G. B. Henning, B. G. Hertz, J. L. Hinton, “Effects of different hypothetical detection mechanisms on the shape of spatial-frequency filters inferred from masking experiments. I. Noise masks,” J. Opt. Soc. Am. 71, 574–581 (1981). [CrossRef] [PubMed]
  21. M. Pavel, G. Sperling, T. Riedl, A. Vanderbeek, “Limits of visual communication: The effect of signal-to-noise ratio on the intelligibility of American Sign Language,” J. Opt. Soc. Am. A 4, 2355–2365 (1987). [CrossRef] [PubMed]
  22. T. R. Riedl, G. Sperling, “Spatial-frequency bands in complex visual stimuli: American Sign Language,” J. Opt. Soc. Am. A 5, 606–616 (1988). [CrossRef] [PubMed]
  23. T. E. Reisbeck, K. R. Gegenfurtner, “Velocity tuned mechanisms in human motion perception,” Invest. Ophthalmol. Visual Sci. Suppl. 38, 376 (1997).
  24. A. Gorea, T. V. Papathomas, “Luminance, color and orientation: local vs. global contrasts in texture segregation,” Invest. Ophthalmol. Visual Sci. Suppl. 39, 649 (1998).
  25. M. D’Zmura, K. Knoblauch, “Spectral bandwidths for the detection of color,” Vision Res. 38, 3117–3128 (1998). [CrossRef]
  26. D. G. Pelli, “Uncertainty explains many aspects of visual contrast detection and discrimination,” J. Opt. Soc. Am. A 2, 1508–1532 (1985). [CrossRef] [PubMed]
  27. A. E. Burgess, B. Colborne, “Visual signal detection. IV. Observer inconsistency,” J. Opt. Soc. Am. A 5, 617–627 (1988). [CrossRef] [PubMed]
  28. A. E. Burgess, “Visual signal detection. III. On Bayesian use of prior knowledge and cross correlation,” J. Opt. Soc. Am. A 2, 1498–1507 (1985). [CrossRef] [PubMed]
  29. B. A. Dosher, Z.-L. Lu, “Attention to location mediated by internal noise reduction,” Invest. Ophthalmol. Visual Sci. Association for Research in Vision and Ophthalmology Suppl. 38, 687 (1997).
  30. Z.-L. Lu, B. A. Dosher, “Characterizing attention mechanisms,” Bull. Psychonom. Soc. 38, 58 (1997).
  31. Z.-L. Lu, B. A. Dosher, “External noise distinguishes mechanisms of attention,” Vision Res. 38, 1183–1198 (1998). [CrossRef] [PubMed]
  32. J. Nachmias, R. V. Sansbury, “Grating contrast: discrimination may be better than detection,” Vision Res. 14, 1039–1042 (1974). [CrossRef] [PubMed]
  33. J. M. Foley, G. E. Legge, “Contrast detection and near-threshold discrimination in human vision,” Vision Res. 21, 1041–1053 (1981). [CrossRef] [PubMed]
  34. C. F. Stromeyer, S. Klein, “Spatial frequency channels in human vision as asymmetric (edge) mechanisms,” Vision Res. 14, 1409–1420 (1974). [CrossRef] [PubMed]
  35. G. E. Legge, J. M. Foley, “Contrast masking in human vision,” J. Opt. Soc. Am. 70, 1458–1471 (1980). [CrossRef] [PubMed]
  36. C. A. Burbeck, D. H. Kelly, “Contrast gain measurements and the transient/sustained dichotomy,” J. Opt. Soc. Am. 71, 1335–1342 (1981). [PubMed]
  37. A. F. Dean, “The relationship between response amplitude and contrast for cat striate cortical neurons,” J. Physiol. (London) 318, 413–427 (1981).
  38. A. M. Derrington, P. Lennie, “Spatial and temporal contrast sensitivities of neurons in lateral geniculate nucleus of macaque,” J. Physiol. (London) 357, 219–240 (1981).
  39. E. Kaplan, R. M. Shapley, “X and Y cells in the lateral geniculate nucleus of macaque monkeys,” J. Physiol. (London) 330, 125–143 (1982).
  40. D. G. Albrecht, D. B. Hamilton, “Striate cortex of monkey and cat: contrast response function,” J. Neurophysiol. 48, 217–237 (1982). [PubMed]
  41. I. Ohzawa, G. Sclar, R. D. Freeman, “Contrast gain control in the cat visual cortex,” Nature (London) 298, 266–268 (1982). [CrossRef]
  42. S. A. Klein, D. M. Levi, “Hyperacuity thresholds of 1 sec: theoretical predictions and empirical validation,” J. Opt. Soc. Am. A 2, 1170–1190 (1985). [CrossRef] [PubMed]
  43. G. Sclar, J. H. Maunsell, P. Lennie, “Coding of image contrast in central visual pathways of the macaque monkey,” Vision Res. 30, 1–10 (1990). [CrossRef] [PubMed]
  44. A. B. Bonds, “Temporal dynamics of contrast gain in single cells of the cat striate cortex,” Visual Neurosci. 6, 239–255 (1991). [CrossRef]
  45. D. G. Albrecht, W. S. Geisler, “Motion selectivity and the contrast-response function of simple cells in the visual cortex,” Visual Neurosci. 7, 531–546 (1991). [CrossRef]
  46. D. J. Heeger, “Normalization of cell responses in cat striate cortex,” Visual Neurosci. 9, 181–197 (1992). [CrossRef]
  47. J. M. Foley, “Human luminance pattern-vision mechanisms: masking experiments require a new model,” J. Opt. Soc. Am. A 11, 1710–1719 (1994). [CrossRef]
  48. Z.-L. Lu, G. Sperling, “Contrast gain control in first- and second-order motion perception,” J. Opt. Soc. Am. A 13, 2305–2318 (1996). [CrossRef]
  49. A. B. Watson, J. A. Solomon, “Model of visual contrast gain control and pattern masking,” J. Opt. Soc. Am. A 14, 2379–2391 (1997). [CrossRef]
  50. M. P. Eckstein, A. J. Ahumada, A. B. Watson, “Visual signal detection in structured backgrounds. II. Effects of contrast gain control, background variations, and white noise,” J. Opt. Soc. Am. A 14, 2406–2419 (1997). [CrossRef]
  51. R. S. Woodworth, H. Schlosberg, Experimental Psychology, 2nd ed. (Holt, Rinehart & Winston, New York, 1954).
  52. D. G. Pelli, L. Zhang, “Accurate control of contrast on microcomputer displays,” Vision Res. 31, 1337–1350 (1991). [CrossRef] [PubMed]
  53. R. E. Fredericksen’ssoftware, “Image manipulation and display library” (1996), was used.
  54. In computing d′, both the LAM [Eq. (3)] and the PTM [Eq. (15)] assume that the noise at the decision stage has a Gaussian distribution. This implies a simple functional relationship between percent correct and d′:Pc=G(d′/2), where G() is cumulative Gaussian. For the LAM, this specifies the cumulative Gaussian with μ=0 (signal contrast) as the form of psychometric functions, while the PTM specifies the form of psychometric function that is cumulative Gaussian with complex functions of cγ as its argument owing to the nonlinearity and multiplicative noise. This provides an alternative way to test the models (see Subsection 5.B). In fact, the cumulative Gaussian specified by the LAM fit the psychometric functions very poorly (see Subsection 5.B). Weibull functions were chosen to fit the psychometric functions for the purpose of estimating (interpolating) three threshold contrast levels. The mean r2 of the Weibull fits to the psychometric data was 0.8818. (A number of other interpolation functions might also have worked. For example, a cumulative Gaussian with guessing correction provided threshold estimates that were essentially identical to the Weibull, r=0.9996, rms difference=0.0019 for a mean threshold of 0.094.) We estimated the bias introduced in the estimation of ratios k1 and k2 when a Weibull was fitted to hypothetical psychometric functions specified by the LAM and when the Weibull was fitted to hypothetical psychometric functions specified by the PTM. In both cases, the biases were of the order of 5% or less, too small to significantly influence the ratio estimates.
  55. W. L. Hays, Statistics, 3rd ed. (CBS College Publishing, New York, 1981).
  56. The standard deviation of each threshold was computed with a resampling method.57 For a given external noise level, the number of correct responses at each of the seven signal contrasts was assumed to have a binomial distribution with a single event probability p equal to the measured percent correct. For each of the seven signal stimulus contrasts in the original psychometric functions at a given external noise level, the number of correct responses at each signal stimulus contrast was independently replaced by a theoretical sample from the corresponding binomial distribution to generate a theoretically resampled psychometric function. Repeating this process 25 times for each external noise condition generated 25 theoretically resampled psychometric functions in every external noise condition. Fitting Weibull functions to each of these psychometric functions and then computing thresholds at 65%, 75%, and 85% performance levels allowed us to estimate the mean and the standard deviations of the parameters in the Weibull function and for the three thresholds.
  57. L. T. Maloney, “Confidence intervals for the parameters of psychometric functions,” Percept. Psychophys. 47, 127–134 (1990). [CrossRef] [PubMed]
  58. Both sets of ratios are closely comparable with those measured for the same observers in the identification task of experiment 1.
  59. B. Zenger, D. Sagi, “Isolating excitatory and inhibitory nonlinear spatial interactions involved in contrast detection,” Vision Res. 36, 2497–2513 (1996). [CrossRef] [PubMed]
  60. R. E. Fredericksen, R. F. Hess, “Estimating multiple temporal mechanisms in human vision,” Vision Res. 38, 1023–1040 (1998). [CrossRef] [PubMed]
  61. In general, the contrast of a signal can be expressed as a function of space and time: S(x, y, t)=cS0(x, y, t) rescaled such that ∭S02(x, y, t)dxdydt=1.0. The contrast of the external noise can be expressed as N(x, y, t)=NextG(x, y, t), where the value of G(x, y, t) at a particular spatiotemporal point (x, y, t) is drawn from a Gaussian distribution with mean 0 and standard deviation 1.0. For a template matching function T(x, y, t), matching the template to a signal-valued stimulus yields the output TS=∭T(x, y, t)S(x, y, t)dxdydt=c∭T(x, y, t)S0(x, y, t)dxdydt; The output of the template matching operation to the external noise is TN=∬T(x, y, t)N(x, y, t)dxdyd t=Next ∭T (x, y, t)G×(x, y, t)dxdydt. For a fixed template and a fixed signal stimulus, TS0=∭T(x, y, t)S0(x, y, t)dxdydt is a constant; TG=∭T(x, y, t)G(x, y, t)dxdydt is a Gaussian random variable with mean 0 and a fixed standard deviation σTG. Because, mathematically, TS and TN can be known only up to a constant, without losing any generality, we set σTG to 1.0. Thus, after template matching, TN=NextG(0, 1),TS=βc, where β=TS0/σTG.
  62. C. R. Carlson, R. W. Klopfenstein, “Spatial-frequency model for hyperacuity,” J. Opt. Soc. Am. A 2, 1747–1751 (1985). [CrossRef]
  63. N. A. MacMillan, C. D. Creelman, Detection Theory: A User’s Guide (Cambridge U. Press, New York, 1991).
  64. In experiment 2, where only one temporal interval contains the signal, the standard deviation of the noise at the decision stage is not identical for the two intervals. This is not an issue in the identification task of experiment 1. To simplify our discussion, we made the approximation of using the standard deviation of the signal-present interval in both situations. In practice, the inaccuracy caused by this approximation is small because noise is dominated by Nadd in the region of low external noise and is dominated by Next in the region of high external noise.
  65. matlab integrated technical computing environment (The MathWorks, Inc., 24 Prime Park Way, Natick, Mass. 01760-1500).
  66. The PTM is a highly nonlinear model with complex dependence on its parameters. The set of analytical constraints on the model parameters developed here allows us to solve for two important parameters independent of the rest in the model. This approach is most advantageous when the solution space of the model parameters is complicated, which is certainly true in the case of the PTM.
  67. The log approximately equates the standard error over large ranges in contrast thresholds, corresponding to weighted least squares, an equivalent to the maximum-likelihood solution for continuous data.
  68. The two models do not yield identical functions, and, furthermore, they partition noise, especially stimulus noise, in a different manner. The EAW model distinguishes between contrast gain control noise associated with structured noise backgrounds in their displays, σcgc, and that as-sociated with additional white-noise variation, σvar. The EAW model assumes the isolation of signal from background noise in the stimulus since contrast-modulated noise depends only on noise, not signal contrast. The PTM considers the power in the stimulus, both signal and external noise, as controlling multiplicative noise.
  69. D. M. Green, J. A. Swets, Signal Detection Theory and Psychophysics (Krieger, Huntington, New York, 1974).
  70. B. Dosher, Z.-L. Lu, “Mechanisms of perceptual learning,” submitted to Vision Res.
  71. F. W. Campbell, J. G. Robson, “Application of Fourier analysis to the visibility of gratings,” J. Physiol. (London) 197, 551–566 (1968).
  72. A. Pantle, R. Sekuler, “Size detecting mechanisms in human vision,” Science 162, 1146–1148 (1969). [CrossRef]
  73. C. Blakemore, F. W. Campbell, “On the existence of neurons in the human visual system selectively sensitive to the orientation and size of retinal images,” J. Physiol. (London) 203, 237–260 (1969).
  74. J. P. Thomas, “Model of the function of receptive fields in human vision,” Psychol. Rev. 77, 121–134 (1970). [CrossRef] [PubMed]
  75. D. H. Hubel, T. N. Wiesel, “Uniformity of monkey striate cortex: a parallel relationship between field size, scatter and magnification factor,” J. Comp. Neurol. 158, 295–306 (1974). [CrossRef] [PubMed]
  76. S. P. Mckee, G. Westheimer, “Improvement in vernier acuity with practice,” Percept. Psychophys. 24, 258–262 (1978). [CrossRef] [PubMed]
  77. H. R. Wilson, J. R. Bergen, “A four mechanism model for threshold spatial vision,” Vision Res. 19, 19–32 (1979). [CrossRef] [PubMed]
  78. N. Graham, “Spatial frequency channels in human vision: detecting edges without edge detectors,” in Visual Coding and Adaptability, C. S. Harris, ed. (Erlbaum, Hillsdale, N.J., 1980), pp. 215–262.
  79. M. A. Webster, R. L. de Valois, “Relationship between spatial-frequency and orientation tuning of striate-cortex cells,” J. Opt. Soc. Am. A 2, 1124–1132 (1985). [CrossRef] [PubMed]

Cited By

Alert me when this paper is cited

OSA is able to provide readers links to articles that cite this paper by participating in CrossRef's Cited-By Linking service. CrossRef includes content from more than 3000 publishers and societies. In addition to listing OSA journal articles that cite this paper, citing articles from other participating publishers will also be listed.

« Previous Article

OSA is a member of CrossRef.

CrossCheck Deposited