English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Tools annual

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Both neurophysiological and psychophysical experiments have pointed out thecrucial role of recurrent and feedback connections to process context-dependentinformation in the early visual cortex. While numerous models have accountedfor feedback effects at either neural or representational level, none of themwere able to bind those two levels of analysis. Is it possible to describefeedback effects at both levels using the same model? We answer this questionby combining Predictive Coding (PC) and Sparse Coding (SC) into a hierarchicaland convolutional framework. In this Sparse Deep Predictive Coding (SDPC)model, the SC component models the internal recurrent processing within eachlayer, and the PC component describes the interactions between layers usingfeedforward and feedback connections. Here, we train a 2-layered SDPC on twodifferent databases of images, and we interpret it as a model of the earlyvisual system (V1 & V2). We first demonstrate that once the training hasconverged, SDPC exhibits oriented and localized receptive fields in V1 and morecomplex features in V2. Second, we analyze the effects of feedback on theneural organization beyond the classical receptive field of V1 neurons usinginteraction maps. These maps are similar to association fields and reflect theGestalt principle of good continuation. We demonstrate that feedback signalsreorganize interaction maps and modulate neural activity to promote contourintegration. Third, we demonstrate at the representational level that the SDPCfeedback connections are able to overcome noise in input images. Therefore, theSDPC captures the association field principle at the neural level which resultsin better disambiguation of blurred images at the representational level.

Sparse Deep Predictive Coding captures contour integration capabilities of the early visual system