The pq-relaxation for the pooling problem can be constructed by applying McCormick envelopes for each of the bilinear terms appearing in the so-called pq-formulation of the pooling problem. This relaxation can be strengthened by using piecewise-linear functions that overand under-estimate each bilinear term. The resulting relaxation can be written as a mixed integer linear programming (MILP) mo...

Bag-of-Words lies at a heart of modern object category recognition systems. After descriptors are extracted from images, they are expressed as vectors representing visual word content, referred to as mid-level features. In this paper, we review a number of techniques for generating mid-level features, including two variants of Soft Assignment, Locality-constrained Linear Coding, and Sparse Codi...

Assuming a convergent projection within a hierarchy of processing stages stimuli from different areas of the receptive .eld project onto the same population of cells. Pooling over space a/ects the representation of individual stimuli, and thus its understanding is crucial for attention and ultimately for object recognition. Since attention, in turn, is likely to modify such spatial pooling by c...

Bilinear pooling has been recently proposed as a feature encoding layer, which can be used after the convolutional layers of a deep network, to improve performance in multiple vision tasks. Different from conventional global average pooling or fully connected layer, bilinear pooling gathers 2nd order information in a translation invariant fashion. However, a serious drawback of this family of p...

Recently, dropout has seen increasing use in deep learning. For deep convolutional neural networks, dropout is known to work well in fully-connected layers. However, its effect in convolutional and pooling layers is still not clear. This paper demonstrates that max-pooling dropout is equivalent to randomly picking activation based on a multinomial distribution at training time. In light of this...