DMCA
Efficient Action Localization with Approximately Normalized Fisher Vectors
| Citations: | 7 - 1 self |
BibTeX
@MISC{Oneata_efficientaction,
author = {Dan Oneata and Jakob Verbeek and Cordelia Schmid},
title = {Efficient Action Localization with Approximately Normalized Fisher Vectors},
year = {}
}
Share
 |
 |
 |
 |
||
OpenURL
Abstract
The Fisher vector (FV) representation is a high-dimensional extension of the popular bag-of-word represen-tation. Transformation of the FV by power and ℓ2 normal-izations has shown to significantly improve its performance, and led to state-of-the-art results for a range of image and video classification and retrieval tasks. These normaliza-tions, however, render the representation non-additive over local descriptors. Combined with its high dimensionality, this makes the FV computationally expensive for the pur-pose of localization tasks. In this paper we present approx-imations to both these normalizations, which yield signifi-cant improvements in the memory and computational costs of the FV when used for localization. Second, we show how these approximations can be used to define upper-bounds on the score function that can be efficiently evaluated, which enables the use of branch-and-bound search as an alterna-tive to exhaustive sliding window search. We present ex-perimental evaluation results on classification and tempo-ral localization of actions in videos. These show that the our approximations lead to a speedup of at least one or-der of magnitude, while maintaining state-of-the-art action recognition and localization performance. 1.
Keyphrases
approximately normalized fisher vector efficient action localization score function present approx-imations window search present ex-perimental evaluation result localization task computational cost tempo-ral localization state-of-the-art result fisher vector branch-and-bound search retrieval task popular bag-of-word represen-tation state-of-the-art action recognition local descriptor high dimensionality video classification localization performance signifi-cant improvement high-dimensional extension
