Generalized Max Pooling

Abstract

State-of-the-art patch-based image representations involve a pooling operation that aggregates statistics computed from local descriptors. Standard pooling operations include sum- and max-pooling. Sum-pooling lacks discriminability because the resulting representation is strongly influenced by frequent yet often uninformative descriptors, but only weakly influenced by rare yet potentially highly-informative ones. Max-pooling equalizes the influence of frequent and rare descriptors but is only applicable to representations that rely on count statistics, such as the bag-of-visual-words (BOV) and its soft- and sparse-coding extensions. We propose a novel pooling mechanism that achieves the same effect as max-pooling but is applicable beyond the BOV and especially to the state-of-the-art Fisher Vector -- hence the name Generalized Max Pooling (GMP). It involves equalizing the similarity between each patch and the pooled representation, which is shown to be equivalent to re-weighting the per-patch statistics. We show on five public image classification benchmarks that the proposed GMP can lead to significant performance gains with respect to heuristic alternatives.

(to appear) CVPR 2014 - IEEE Conference on Computer Vision & Pattern Recognition (2014)

Cited by in corpus (16)

• Object-Part Attention Model for Fine-grained Image Classification
• Deep Image: Scaling up Image Recognition
• Good Practice in CNN Feature Transfer
• Neural Activation Constellations: Unsupervised Part Model Discovery with Convolutional Networks
• Metric Learning with Adaptive Density Discrimination
• A comparison of dense region detectors for image search and fine-grained classification
• SIFT Meets CNN: A Decade Survey of Instance Retrieval
• client2vec: Towards Systematic Baselines for Banking Applications
• Simultaneous Feature Aggregating and Hashing for Large-scale Image Search
• Multipartite Pooling for Deep Convolutional Neural Networks
• From Selective Deep Convolutional Features to Compact Binary Representations for Image Retrieval
• Embedding based on function approximation for large scale image search
• Fast Spectral Ranking for Similarity Search
• Colour Constancy: Biologically-inspired Contrast Variant Pooling Mechanism
• Convolution in Convolution for Network in Network
• Vision Recognition using Discriminant Sparse Optimization Learning