Visual Attention Emerges from Recurrent Sparse Reconstruction. (arXiv:2204.10962v2 [cs.CV] UPDATED)

Visual attention helps achieve robust perception under noise, corruption, and
distribution shifts in human vision, which are areas where modern neural
networks still fall short. We present VARS, Visual Attention from Recurrent
Sparse reconstruction, a new attention formulation built on two prominent
features of the human visual attention mechanism: recurrency and sparsity.
Related features are grouped together via recurrent connections between
neurons, with salient objects emerging via sparse regularization. VARS adopts
an attractor network with recurrent connections that converges toward …

arxiv attention cv visual attention