Sensory systems are faced with an essentially infinite number of possible environmental events but have limited processing resources. Posed with this challenge, it makes sense to allocate these resources to prioritize the discrimination of events with the most behavioral relevance. Here, we asked if such relevance is reflected in the processing and perception of motion. We compared human performance on a rapid motion direction discrimination task, including monocular and binocular viewing. In particular, we determined sensitivity and bias for a binocular motion-in-depth (three-dimensional; 3D) stimulus and for its constituent monocular (two-dimensional; 2D) signals over a broad range of speeds. Consistent with prior work, we found that binocular 3D sensitivity was lower than monocular sensitivity for all speeds. Although overall sensitivity was worse for 3D discrimination, we found that the transformation from 2D to 3D motion processing also incorporated a pattern of potentially advantageous biases. One such bias is reflected by a criterion shift that occurs at the level of 3D motion processing and results in an increased hit rate for motion toward the head. We also observed an increase in sensitivity for 3D motion trajectories presented on crossed rather than uncrossed disparity pedestals, privileging motion trajectories closer to the observer. We used these measurements to determine the range of real-world trajectories for which rapid 3D motion discrimination is most useful. These results suggest that the neural mechanisms that underlie motion perception privilege behaviorally relevant motion and provide insights into the nature of human motion sensitivity in the real world.

People make surprising but reliable perceptual errors. Here, we provide a unified explanation for systematic errors in the perception of three-dimensional (3-D) motion. To do so, we characterized the binocular retinal motion signals produced by objects moving through arbitrary locations in 3-D. Next, we developed a Bayesian model, treating 3-D motion perception as optimal inference given sensory noise in the measurement of retinal motion. The model predicts a set of systematic perceptual errors, which depend on stimulus distance, contrast, and eccentricity. We then used a virtual-reality headset as well as a standard 3-D desktop stereoscopic display to test these predictions in a series of perceptual experiments. As predicted, we found evidence that errors in 3-D motion perception depend on the contrast, viewing distance, and eccentricity of a stimulus. These errors include a lateral bias in perceived motion direction and a surprising tendency to misreport approaching motion as receding and vice versa. In sum, we present a Bayesian model that provides a parsimonious account for a range of systematic misperceptions of motion in naturalistic environments.

People make surprising but reliable perceptual errors. Here, we provide a unified explanation for systematic errors in the perception of three-dimensional (3-D) motion. To do so, we characterized the binocular retinal motion signals produced by objects moving through arbitrary locations in 3-D. Next, we developed a Bayesian model, treating 3-D motion perception as optimal inference given sensory noise in the measurement of retinal motion. The model predicts a set of systematic perceptual errors, which depend on stimulus distance, contrast, and eccentricity. We then used a virtual-reality headset as well as a standard 3-D desktop stereoscopic display to test these predictions in a series of perceptual experiments. As predicted, we found evidence that errors in 3-D motion perception depend on the contrast, viewing distance, and eccentricity of a stimulus. These errors include a lateral bias in perceived motion direction and a surprising tendency to misreport approaching motion as receding and vice versa. In sum, we present a Bayesian model that provides a parsimonious account for a range of systematic misperceptions of motion in naturalistic environments.

Metacognitive awareness-the ability to know that one is having a particular experience-is thought to guide optimal behavior, but its neural bases continue to be the subject of vigorous debate. Prior work has identified correlations between perceptual metacognitive ability and the structure and function of lateral prefrontal cortex (LPFC); however, evidence for a causal role of this region in promoting metacognition is controversial. Moreover, whether LPFC function promotes metacognitive awareness of perceptual and emotional features of complex, yet ubiquitous face stimuli is unknown. Here, using model-based analyses following a causal intervention to LPFC in humans, we demonstrate that LPFC function promotes metacognitive awareness of the orientation of faces-although not of their emotional expressions. Collectively, these data support the causal involvement of the prefrontal cortex in metacognitive awareness, and indicate that the role of LPFC in metacognition encompasses perceptual experiences of naturalistic social stimuli.

A central question in neuroscience is how sensory inputs are transformed into percepts. At this point, it is clear that this process is strongly influenced by prior knowledge of the sensory environment. Bayesian ideal observer models provide a useful link between data and theory that can help researchers evaluate how prior knowledge is represented and integrated with incoming sensory information. However, the statistical prior employed by a Bayesian observer cannot be measured directly, and must instead be inferred from behavioral measurements. Here, we review the general problem of inferring priors from psychophysical data, and the simple solution that follows from assuming a prior that is a Gaussian probability distribution. As our understanding of sensory processing advances, however, there is an increasing need for methods to flexibly recover the shape of Bayesian priors that are not well approximated by elementary functions. To address this issue, we describe a novel approach that applies to arbitrary prior shapes, which we parameterize using mixtures of Gaussian distributions. After incorporating a simple approximation, this method produces an analytical solution for psychophysical quantities that can be numerically optimized to recover the shapes of Bayesian priors. This approach offers advantages in flexibility, while still providing an analytical framework for many scenarios. We provide a MATLAB toolbox implementing key computations described herein.