Optimal Local Estimates of Visual Motion in a Natural Environment

Many organisms use visual signals to estimate motion, and these estimates typically are biased. Here, we ask whether these biases may reflect physical rather than biological limitations. Using a camera-gyroscope system, we sample the joint distribution of images and rotational motions in a natural e...

Full description

Saved in:
Bibliographic Details
Published in:Physical review letters 2021-01, Vol.126 (1), p.018101-018101, Article 018101
Main Authors: Sinha, Shiva R, Bialek, William, van Steveninck, Rob R de Ruyter
Format: Article
Language:English
Subjects:
Animals
Calliphoridae - physiology
Environment
Models, Biological
Motion Perception - physiology
Photography
Signal processing
Visual signals
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Many organisms use visual signals to estimate motion, and these estimates typically are biased. Here, we ask whether these biases may reflect physical rather than biological limitations. Using a camera-gyroscope system, we sample the joint distribution of images and rotational motions in a natural environment, and from this distribution we construct the optimal estimator of velocity based on local image intensities. Over most of the natural dynamic range, this estimator exhibits the biases observed in neural and behavioral responses. Thus, imputed errors in sensory processing may represent an optimal response to the physical signals sampled from the environment.
ISSN:0031-9007
1079-7114
DOI:10.1103/PhysRevLett.126.018101