A Neural Population Model for Visual Pattern Detection
A Neural Population Model for Visual Pattern Detection
复制标题
DOI：
10.1037/a0033136
复制
发表时间：
2013-07-01
影响因子：
5.4
通讯作者：
Wichmann, Felix A.
中科院分区：
文献类型：
--
作者：
Goris, Robbe L. T.;Putzeys, Tom;Wichmann, Felix A.
模式检测是现代视觉科学的基石。大约半个世纪前，心理物理学家提倡一个定量的理论框架，该框架将视觉模式检测与其神经生理学的基础联系起来。在这一理论中，主要视觉皮层中的神经元构成线性和独立的视觉通道，其输出通过简单读出机制与检测任务中的选择行为相关联。事实证明，该模型在计算阈值愿景方面非常成功。然而，从根本上讲，这是矛盾的，目前对模式视觉的神经生理基础知识的了解。此外，该模型中提出的原则未能推广到除检测以外的其他视野或感知任务。我们提出了一种替代性检测理论，在该理论中，感知决策是由神经生理启发的原代视觉皮层中人群活性模型的最大样本解码而发展的。我们证明了该理论解释了广泛的经典检测结果。有了一组参数，我们的模型可以说明几个求和，适应和不确定性效应，从而为大量的心理物理文献提供了有关模式检测的新理论解释。
Pattern detection is the bedrock of modern vision science. Nearly half a century ago, psychophysicists advocated a quantitative theoretical framework that connected visual pattern detection with its neurophysiological underpinnings. In this theory, neurons in primary visual cortex constitute linear and independent visual channels whose output is linked to choice behavior in detection tasks via simple read-out mechanisms. This model has proven remarkably successful in accounting for threshold vision. It is fundamentally at odds, however, with current knowledge about the neurophysiological underpinnings of pattern vision. In addition, the principles put forward in the model fail to generalize to suprathreshold vision or perceptual tasks other than detection. We propose an alternative theory of detection in which perceptual decisions develop from maximum-likelihood decoding of a neurophysiologically inspired model of population activity in primary visual cortex. We demonstrate that this theory explains a broad range of classic detection results. With a single set of parameters, our model can account for several summation, adaptation, and uncertainty effects, thereby offering a new theoretical interpretation for the vast psychophysical literature on pattern detection.