Bayesian inference models with efficient encoding constraints in visual perception: Prospects for their application

The computational model built upon the efficient coding hypothesis and Bayesian inference theory has become an important theoretical framework in the field of visual cognition research. This model not only effectively explains various phenomena related to visual perception, but also provides a unified computational framework for systematically understanding the entire process of visual information processing. This paper provides a systematic review of the theoretical foundations of this model and its applications in visual cognition research. Based on current research, it is possible to further validate the neural physiological basis of this model by combining it with neuroelectrophysiological techniques. Additionally, this model can be extended to other areas of sensory perception, advanced social cognition, and specific populations.