A long-standing mystery to neuroscientists is how perceptual systems rapidly and effortlessly compute a vivid and veridical representation of the external world from the noisy and ambiguous input furnished by our sensory systems. Most neuroscientists now agree that problem is in part solved by the fact that the brain does not process all incoming sensory information anew, but generates a model of the world from past experience, and regularly updates this model from current sensory data. This classic idea has recently been well formulised within the modern framework of Generative Bayesian Inference, but the mechanisms remain poorly understood, both from a functional and neural point of view.
The objectives of GenPercept were to demonstrate the fundamental importance of generative processes in perception, showing how they lead to efficient perceptual processes. It proposed to characterise quantitatively their functional role, then go on to explore the underlying neural mechanisms, exploiting state-of-the-art psychophysical, EEG, imaging and pupillometry techniques. A major goal was to explore the innovative idea that neural oscillations reflect reverberations in the propagation of generative prediction and error signals. All of the objectives have been realized.
The major importance of the study is to science, increasing our understanding of how we perceive and interact with the world, and how past perceptual experience influences what we perceive. There are no immediate clinical or societal benefits, but in the long term, increasing basic understanding in this area can benefit many fields, from ophthalmology to neurology, and even computer science and robotics.
