Recovering real-world images from single-scale boundaries with a novel filling-in architecture

Abstract

Filling-in models were successful in predicting psychophysical data for brightness perception. Nevertheless, their suitability for real-world image processing has never been examined. A unified architecture for both predicting psychophysical data and real-world image processing would constitute a powerful theory for early visual information processing. As a first contribution of the present paper, we identified three principal problems with current filling-in architectures, which hamper the goal of having such a unified architecture. To overcome these problems we propose an advance to filling-in theory, called BEATS filling-in, which is based on a novel nonlinear diffusion operator. BEATS filling-in furthermore introduces novel boundary structures. We compare, by means of simulation studies with real-world images, the performance of BEATS filling-in with the recently proposed confidence-based filling-in. As a second contribution we propose a novel mechanism for encoding luminance information in contrast responses ('multiplex contrasts'), which is based on recent neurophysiological findings. Again, by simulations, we show that 'multiplex contrasts' at a single, high-resolution filter scale are sufficient for recovering absolute luminance levels. Hence, 'multiplex contrasts' represent a novel theory addressing how the brain encodes and decodes luminance information. q

Local to global normalization dynamic by nonlinear local interactions

“…As a first step into that direction, a simple adaptation algorithm based on the maximisation of entropy was proposed (section III D): the dynamic is frozen as soon as a maximum of entropy is reached, and the output is then fed back as new input to the dynamic normalization network. As a further improvement, the diffusion operators could be modified such that activity exchange between two cells is blocked for sufficiently large activity gradients [35]. Doing so would possibly prevent in figure 9 (first and second row) the global maximum from spreading between tiles, and would normalize each tile independently, such that ideally a dynamic similar to the bottom row in figure 9 is produced.…”

“…Doing so would possibly prevent in figure 9 (first and second row) the global maximum from spreading between tiles, and would normalize each tile independently, such that ideally a dynamic similar to the bottom row in figure 9 is produced. Systems based on pseudo-diffusion have already turned out to be of utility for a variety of purposes in image processing (for implementing filling-in mechanisms, or winner-takes all inhibition, see [35]). Pseudo-diffusion Initially, cells in the dynamic normalization layer have small activities and concentrate in the first histogram bins (upper left corner; the first 14 time steps were dropped for visualization reasons).…”

Predictive coding as a unifying principle for explaining a broad range of brightness phenomena

“…To this end, he relies heavily on the max-operator, which he justified 139 with a dendritic circuit proposal. In comparison, [29] used a modified diffusion operator, 140 which could be efficiently implemented (both physiologically and computationally) with 141 rectifying gap-junctions or rectifying dendro-dendritic connections [17,29]. For filling-in, 142 Domijan's used a luminance sensitive pathway.…”

“…For filling-in, 142 Domijan's used a luminance sensitive pathway. Luminance-modulated contrast 143 responses are computed with an unbalanced center-surround kernel, similar to [30], but 144 see also [17] for a different way of computing multiplexed contrasts). In addition, [28] 145 computed BCS activity by first deriving a local boundary map, where the loss of activity 146 at junctions and corners was corrected.…”

Depth-resolved ultra-high field fMRI reveals feedback contributions to surface motion perception

“…Surface perception is thought to interact tightly with mechanisms of contour reconstruc-tion. A number of computational models of surface perception (Grossberg, 1987a(Grossberg, , 1987b; see also Keil, Cristóbal, Hansen, & Neumann, 2005) have proposed that diffusion-like spreading in a surface feature system is contained within proper retinotopic bounds by local inhibition delivered by boundary representations. Neurophysiological observations of contour-related responses in V2 (von der Heydt, Peterhans, & Baumgartner, 1984) and in V1 (Grosof, Shapley, & Hawken, 1993) and surface related responses in V1, V2 and V3 (De Weerd et al, 1995;Huang & Paradiso, 2008) have emphasized the role of early visual areas in this interaction between surface and contour processing.…”