Backpropagation Training for Fisher Vectors within Neural Networks

TL;DR

This paper introduces a neural network framework that enables end-to-end training of Fisher Vectors, allowing joint learning of feature representations, FV parameters, and classifiers, leading to improved visual recognition performance.

Contribution

It presents a novel method to embed Fisher Vectors into neural networks for joint optimization via backpropagation, which was not previously possible.

Findings

Improved performance on Pascal VOC 2007 with multi-GPU training.

Demonstrated embedding of FV into neural networks at arbitrary positions.

Achieved end-to-end training of FV within deep architectures.

Abstract

Fisher-Vectors (FV) encode higher-order statistics of a set of multiple local descriptors like SIFT features. They already show good performance in combination with shallow learning architectures on visual recognitions tasks. Current methods using FV as a feature descriptor in deep architectures assume that all original input features are static. We propose a framework to jointly learn the representation of original features, FV parameters and parameters of the classifier in the style of traditional neural networks. Our proof of concept implementation improves the performance of FV on the Pascal Voc 2007 challenge in a multi-GPU setting in comparison to a default SVM setting. We demonstrate that FV can be embedded into neural networks at arbitrary positions, allowing end-to-end training with back-propagation.