Singh Vijay, Tchernookov Martin, Butterfield Rebecca, Nemenman Ilya
Department of Physics, Emory University, Atlanta, GA, United States of America.
Department of Biology, Emory University, Atlanta, GA, United States of America.
PLoS One. 2014 Oct 17;9(10):e108991. doi: 10.1371/journal.pone.0108991. eCollection 2014.
We aim to build the simplest possible model capable of detecting long, noisy contours in a cluttered visual scene. For this, we model the neural dynamics in the primate primary visual cortex in terms of a continuous director field that describes the average rate and the average orientational preference of active neurons at a particular point in the cortex. We then use a linear-nonlinear dynamical model with long range connectivity patterns to enforce long-range statistical context present in the analyzed images. The resulting model has substantially fewer degrees of freedom than traditional models, and yet it can distinguish large contiguous objects from the background clutter by suppressing the clutter and by filling-in occluded elements of object contours. This results in high-precision, high-recall detection of large objects in cluttered scenes. Parenthetically, our model has a direct correspondence with the Landau-de Gennes theory of nematic liquid crystal in two dimensions.
我们旨在构建一个尽可能简单的模型，该模型能够在杂乱的视觉场景中检测出长的、有噪声的轮廓。为此，我们根据一个连续的指向场对灵长类动物初级视觉皮层中的神经动力学进行建模，该指向场描述了皮层中特定点处活跃神经元的平均发放率和平均方向偏好。然后，我们使用具有长程连接模式的线性 - 非线性动力学模型来强化分析图像中存在的长程统计背景。与传统模型相比，所得模型的自由度大幅减少，然而它能够通过抑制背景杂波并填充物体轮廓中被遮挡的元素，将大的连续物体与背景杂波区分开来。这导致在杂乱场景中对大物体进行高精度、高召回率的检测。顺便提一下，我们的模型与二维向列型液晶的朗道 - 德热纳理论有直接对应关系。
