Natural Image Coding in V1: How Much Use Is Orientation Selectivity?

Orientation selectivity is the most striking feature of simple cell coding in V1 that has been shown to emerge from the reduction of higher-order correlations in natural images in a large variety of statistical image models. The most parsimonious one among these models is linear Independent Component Analysis (ICA), whereas second-order decorrelation transformations such as Principal Component Analysis (PCA) do not yield oriented filters. Because of this finding, it has been suggested that the emergence of orientation selectivity may be explained by higher-order redundancy reduction. To assess the tenability of this hypothesis, it is an important empirical question how much more redundancy can be removed with ICA in comparison to PCA or other second-order decorrelation methods. Although some previous studies have concluded that the amount of higher-order correlation in natural images is generally insignificant, other studies reported an extra gain for ICA of more than 100%. A consistent conclusion about the role of higher-order correlations in natural images can be reached only by the development of reliable quantitative evaluation methods. Here, we present a very careful and comprehensive analysis using three evaluation criteria related to redundancy reduction: In addition to the multi-information and the average log-loss, we compute complete rate–distortion curves for ICA in comparison with PCA. Without exception, we find that the advantage of the ICA filters is small. At the same time, we show that a simple spherically symmetric distribution with only two parameters can fit the data significantly better than the probabilistic model underlying ICA. This finding suggests that, although the amount of higher-order correlation in natural images can in fact be significant, the feature of orientation selectivity does not yield a large contribution to redundancy reduction within the linear filter bank models of V1 simple cells.Author Summary: Since the Nobel Prize winning work of Hubel and Wiesel it has been known that orientation selectivity is an important feature of simple cells in the primary visual cortex. The standard description of this stage of visual processing is that of a linear filter bank where each neuron responds to an oriented edge at a certain location within the visual field. From a vision scientist's point of view, we would like to understand why an orientation selective filter bank provides a useful image representation. Several previous studies have shown that orientation selectivity arises when the individual filter shapes are optimized according to the statistics of natural images. Here, we investigate quantitatively how critical the feature of orientation selectivity is for this optimization. We find that there is a large range of non-oriented filter shapes that perform nearly as well as the optimal orientation selective filters. We conclude that the standard filter bank model is not suitable to reveal a strong link between orientation selectivity and the statistics of natural images. Thus, to understand the role of orientation selectivity in the primary visual cortex, we will have to develop more sophisticated, nonlinear models of natural images.