Abstract
Understanding how nervous systems exploit task-relevant properties of sensory stimuli to perform natural tasks is fundamental to the study of perceptual systems. However, there are few formal methods for determining which stimulus properties are most useful for a given natural task. As a consequence, it is difficult to develop principled models for how to compute task-relevant latent variables from natural signals, and it is difficult to evaluate descriptive models fit to neural response. Accuracy maximization analysis (AMA) is a recently developed Bayesian method for finding the optimal task-specific filters (receptive fields). Here, we introduce AMA–Gauss, a new faster form of AMA that incorporates the assumption that the class-conditional filter responses are Gaussian distributed. Then, we use AMA–Gauss to show that its assumptions are justified for two fundamental visual tasks: retinal speed estimation and binocular disparity estimation. Next, we show that AMA–Gauss has striking formal similarities to popular quadratic models of neural response: the energy model and the generalized quadratic model (GQM). Together, these developments deepen our understanding of why the energy model of neural response have proven useful, improve our ability to evaluate results from subunit model fits to neural data, and should help accelerate psychophysics and neuroscience research with natural stimuli.
Highlights
Perceptual systems capture and process sensory stimuli to obtain information about behaviorally relevant properties of the environment
We use Accuracy maximization analysis (AMA)–Gauss1 to find the receptive fields that are optimal for estimating speed and disparity from local patches of natural images
Last, we establish the formal relationship between AMA–Gauss and the generalized quadratic model (GQM), a recently developed method for neural systems identification
Summary
Perceptual systems capture and process sensory stimuli to obtain information about behaviorally relevant properties of the environment. Perceptual and neural processing in particular tasks is driven by sets of features that occupy a lower dimensional space (i.e., can be described more compactly) than the stimuli These considerations have motivated perception and neuroscience researchers to develop methods for dimensionality reduction that characterize the statistical properties of proximal stimuli, that describe the responses of neurons to those stimuli, and that specify how those responses could be decoded (Bell & Sejnowski, 1997; Cook & Forzani, 2009; Cook, Forzani, & Yao, 2010; Hotelling, 1933; Lewicki, 2002; McFarland, Cui, & Butts, 2013; Olshausen & Field, 1996; Pagan, Simoncelli, & Rust, 2016; Park, Archer, Priebe, & Pillow, 2013; Ruderman & Bialek, 1994; Rust, Schwartz, Movshon, & Simoncelli, 2005; Schwartz, Pillow, Rust, & Simoncelli, 2006; Tipping & Bishop, 1999; Vintch, Movshon, & Simoncelli, 2015). It is likely to help accelerate the development of normative models of other tasks for which the energy model has provided a useful description
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
