Active Efficient Coding Explains the Development of Binocular Vision and its Failure in Amblyopia

Abstract

The development of vision during the first months of life is an active process that comprises the learning of appropriate neural representations and the learning of accurate eye movements. While it has long been suspected that the two learning processes are coupled, there is still no widely accepted theoretical framework describing this joint development. Here we propose a computational model of the development of active binocular vision to fill this gap. The model is based on a new formulation of the Active Efficient Coding theory, which proposes that eye movements, as well as stimulus encoding, are jointly adapted to maximize the overall coding efficiency. Under healthy conditions, the model self-calibrates to perform accurate vergence and accommodation eye movements. It exploits disparity cues to deduce the direction of defocus, which leads to coordinated vergence and accommodation responses. In a simulated anisometropic case, where the refraction power of the two eyes differs, an amblyopia-like state develops, in which the foveal region of one eye is suppressed due to inputs from the other eye. After correcting for refractive errors, the model can only reach healthy performance levels if receptive fields are still plastic, in line with findings on a critical period for binocular vision development. Overall, our model offers a unifying conceptual framework for understanding the development of binocular vision under healthy conditions and in amblyopia. efficient coding | active perception | amblyopia | vergence | accommodation

Unsupervised learning of haptic material properties

“…Specific to exploration speed, it is shown that when participants were asked to haptically discriminate spatial frequencies of gratings, low performing participants could improve their performance by adjusting the scanning velocity to the one used by better performing participants (Gamzu & Ahissar, 2001). In fact, the statistics of the sensory input are shaped by exploratory behavior, which therefore may contribute to efficient encoding, consistent with the recently proposed "active efficient coding" theory (Eckmann et al, 2020).…”

“…Receptive field properties in the early visual pathway (Atick & Redlich, 1992; Olshausen & Field, 1996;1997) as well as the tuning properties of auditory nerve fibers (Lewicki, 2002; Smith & Lewicki, 2006) can emerge by efficiently encoding natural images or sounds respectively. Recently, it was shown that efficient coding could also explain the simultaneous development of vergence and accommodation as a result of maximizing coding efficiency of the retinal signals (Eckmann, Klimmasch, Shi, & Triesch, 2020). There is currently a lot of interest whether higher level representations can also be learned by efficient encoding of the retinal images (Fleming & Storrs, 2019).…”

Predictive coding is a consequence of energy efficiency in recurrent neural networks

“…In search for such principles, we take inspiration from the organism’s limited energy budget. Expanding on previous work on efficient coding and sparse coding (Barlow et al, 1961; Bell and Sejnowski, 1995; Olshausen and Field, 1996; Bialek et al, 2006; Berkes and Wiskott, 2005; Chalk et al, 2018; Eckmann et al, 2020), we subject recurrent neural networks (RNNs) to predictable sequences of visual input and optimise their synaptic weights to minimise what constitutes the largest source of energy consumption in biological systems: action potential generation and synaptic transmission (Sengupta et al, 2010). We then test the resulting networks for phenomena typically associated with predictive coding.…”

Active head rolls enhance sonar-based auditory localization performance

“…Our model is based on the Active Efficient Coding (AEC) framework, an extension of classic efficient coding ideas [ 14 , 15 ] to active perception. Previous AEC models have been restricted to the visual domain, where they have explained the self-calibration of active binocular vision, active motion vision, the control of accommodation and torsional eye movements and combinations thereof [ 16 , 17 , 28 , 29 ]. Moreover, AEC models have been developed to learn eye hand coordination tasks such as autonomous tracking of a robot arm [ 30 , 31 ].…”