Second-order Democratic Aggregation

08/22/2018
by   Tsung-Yu Lin, et al.
0

Aggregated second-order features extracted from deep convolutional networks have been shown to be effective for texture generation, fine-grained recognition, material classification, and scene understanding. In this paper, we study a class of orderless aggregation functions designed to minimize interference or equalize contributions in the context of second-order features and we show that they can be computed just as efficiently as their first-order counterparts and they have favorable properties over aggregation by summation. Another line of work has shown that matrix power normalization after aggregation can significantly improve the generalization of second-order representations. We show that matrix power normalization implicitly equalizes contributions during aggregation thus establishing a connection between matrix normalization techniques and prior work on minimizing interference. Based on the analysis we present γ-democratic aggregators that interpolate between sum (γ=1) and democratic pooling (γ=0) outperforming both on several classification tasks. Moreover, unlike power normalization, the γ-democratic aggregations can be computed in a low dimensional space by sketching that allows the use of very high-dimensional second-order features. This results in a state-of-the-art performance on several datasets.

READ FULL TEXT

page 1

page 2

page 3

page 4

research
01/23/2018

Statistically Motivated Second Order Pooling

Second-order pooling, a.k.a. bilinear pooling, has proven effective for ...
research
03/30/2020

Multi-Objective Matrix Normalization for Fine-grained Visual Recognition

Bilinear pooling achieves great success in fine-grained visual recogniti...
research
10/07/2013

Discriminative Features via Generalized Eigenvectors

Representing examples in a way that is compatible with the underlying cl...
research
07/21/2017

Improved Bilinear Pooling with CNNs

Bilinear pooling of Convolutional Neural Network (CNN) features [22, 23]...
research
12/27/2020

Power Normalizations in Fine-grained Image, Few-shot Image and Graph Classification

Power Normalizations (PN) are useful non-linear operators which tackle f...
research
11/29/2018

Global Second-order Pooling Convolutional Networks

Deep Convolutional Networks (ConvNets) are fundamental to, besides large...
research
06/17/2016

Most central or least central? How much modeling decisions influence a node's centrality ranking in multiplex networks

To understand a node's centrality in a multiplex network, its centrality...

Please sign up or login with your details

Forgot password? Click here to reset