Limits...
Nonlinear circuits for naturalistic visual motion estimation.

Fitzgerald JE, Clark DA - Elife (2015)

Bottom Line: Furthermore, a diversity of inputs to motion detecting neurons can provide access to more complex higher-order correlations.Collectively, these results illustrate how non-canonical computations improve motion estimation with naturalistic inputs.This argues that the complexity of the fly's motion computations, implemented in its elaborate circuits, represents a valuable feature of its visual motion estimator.

View Article: PubMed Central - PubMed

Affiliation: Center for Brain Science, Harvard University, Cambridge, United States.

ABSTRACT
Many animals use visual signals to estimate motion. Canonical models suppose that animals estimate motion by cross-correlating pairs of spatiotemporally separated visual signals, but recent experiments indicate that humans and flies perceive motion from higher-order correlations that signify motion in natural environments. Here we show how biologically plausible processing motifs in neural circuits could be tuned to extract this information. We emphasize how known aspects of Drosophila's visual circuitry could embody this tuning and predict fly behavior. We find that segregating motion signals into ON/OFF channels can enhance estimation accuracy by accounting for natural light/dark asymmetries. Furthermore, a diversity of inputs to motion detecting neurons can provide access to more complex higher-order correlations. Collectively, these results illustrate how non-canonical computations improve motion estimation with naturalistic inputs. This argues that the complexity of the fly's motion computations, implemented in its elaborate circuits, represents a valuable feature of its visual motion estimator.

Show MeSH
The performance of the non-multiplicative nonlinearity model is plotted against the order of the fitted polynomial.With only zeroth or first-order terms, the model cannot predict motion. With second-order terms, it can perform slightly better than the HRC (Appendix 10). The biggest performance increase occurred when third-order terms were included, and the fourth-order terms also improved performance.DOI:http://dx.doi.org/10.7554/eLife.09123.012
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4663970&req=5

fig4s2: The performance of the non-multiplicative nonlinearity model is plotted against the order of the fitted polynomial.With only zeroth or first-order terms, the model cannot predict motion. With second-order terms, it can perform slightly better than the HRC (Appendix 10). The biggest performance increase occurred when third-order terms were included, and the fourth-order terms also improved performance.DOI:http://dx.doi.org/10.7554/eLife.09123.012

Mentions: Having introduced the rationale behind the non-multiplicative, unrestricted, and extra input nonlinearity models, it is straightforward to examine their performance as motion estimators. First note that the polynomial non-multiplicative nonlinearity model was a better motion estimator (Figure 4E) than the weighted 4-quadrant model (Figure 3C). This implies that some useful signatures of naturalistic motion are not made accessible by simply segregating ON and OFF motion signals. Interestingly, this performance improvement is largely due to 3-point correlations, and models that exclude fourth-order polynomial terms still outperform the weighted 4-quadrant model (Figure 4—figure supplement 2). Third-order correlations are only useful for motion estimation because of light–dark asymmetries in natural stimulus statistics (Fitzgerald et al., 2011; Clark et al., 2014), so this result implies that ON/OFF segregation provides an imperfect way to account for the complexity of light–dark asymmetries found in the natural world. The non-multiplicative nonlinearity model also made novel use of low-order correlations to improve its motion estimate (Appendix 10).


Nonlinear circuits for naturalistic visual motion estimation.

Fitzgerald JE, Clark DA - Elife (2015)

The performance of the non-multiplicative nonlinearity model is plotted against the order of the fitted polynomial.With only zeroth or first-order terms, the model cannot predict motion. With second-order terms, it can perform slightly better than the HRC (Appendix 10). The biggest performance increase occurred when third-order terms were included, and the fourth-order terms also improved performance.DOI:http://dx.doi.org/10.7554/eLife.09123.012
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4663970&req=5

fig4s2: The performance of the non-multiplicative nonlinearity model is plotted against the order of the fitted polynomial.With only zeroth or first-order terms, the model cannot predict motion. With second-order terms, it can perform slightly better than the HRC (Appendix 10). The biggest performance increase occurred when third-order terms were included, and the fourth-order terms also improved performance.DOI:http://dx.doi.org/10.7554/eLife.09123.012
Mentions: Having introduced the rationale behind the non-multiplicative, unrestricted, and extra input nonlinearity models, it is straightforward to examine their performance as motion estimators. First note that the polynomial non-multiplicative nonlinearity model was a better motion estimator (Figure 4E) than the weighted 4-quadrant model (Figure 3C). This implies that some useful signatures of naturalistic motion are not made accessible by simply segregating ON and OFF motion signals. Interestingly, this performance improvement is largely due to 3-point correlations, and models that exclude fourth-order polynomial terms still outperform the weighted 4-quadrant model (Figure 4—figure supplement 2). Third-order correlations are only useful for motion estimation because of light–dark asymmetries in natural stimulus statistics (Fitzgerald et al., 2011; Clark et al., 2014), so this result implies that ON/OFF segregation provides an imperfect way to account for the complexity of light–dark asymmetries found in the natural world. The non-multiplicative nonlinearity model also made novel use of low-order correlations to improve its motion estimate (Appendix 10).

Bottom Line: Furthermore, a diversity of inputs to motion detecting neurons can provide access to more complex higher-order correlations.Collectively, these results illustrate how non-canonical computations improve motion estimation with naturalistic inputs.This argues that the complexity of the fly's motion computations, implemented in its elaborate circuits, represents a valuable feature of its visual motion estimator.

View Article: PubMed Central - PubMed

Affiliation: Center for Brain Science, Harvard University, Cambridge, United States.

ABSTRACT
Many animals use visual signals to estimate motion. Canonical models suppose that animals estimate motion by cross-correlating pairs of spatiotemporally separated visual signals, but recent experiments indicate that humans and flies perceive motion from higher-order correlations that signify motion in natural environments. Here we show how biologically plausible processing motifs in neural circuits could be tuned to extract this information. We emphasize how known aspects of Drosophila's visual circuitry could embody this tuning and predict fly behavior. We find that segregating motion signals into ON/OFF channels can enhance estimation accuracy by accounting for natural light/dark asymmetries. Furthermore, a diversity of inputs to motion detecting neurons can provide access to more complex higher-order correlations. Collectively, these results illustrate how non-canonical computations improve motion estimation with naturalistic inputs. This argues that the complexity of the fly's motion computations, implemented in its elaborate circuits, represents a valuable feature of its visual motion estimator.

Show MeSH