Comparison of mid-level feature coding approaches and pooling strategies in visual concept detection

Koniusz, P, Yan, F and Mikolajczyk, K (2013) Comparison of mid-level feature coding approaches and pooling strategies in visual concept detection Computer Vision and Image Understanding, 117 (5). pp. 479-492.

Text elsarticle-template-single.pdf Restricted to Repository staff only Download (1MB)

Abstract

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 Coding. We also isolate the underlying properties that affect their performance. Moreover, we investigate various pooling methods that aggregate mid-level features into vectors representing images. Average pooling, Max-pooling, and a family of likelihood inspired pooling strategies are scrutinised. We demonstrate how both coding schemes and pooling methods interact with each other. We generalise the investigated pooling methods to account for the descriptor interdependence and introduce an intuitive concept of improved pooling. We also propose a coding-related improvement to increase its speed. Lastly, state-of-the-art performance in classification is demonstrated on Caltech101, Flower17, and ImageCLEF11 datasets. © 2012 Elsevier Inc. All rights reserved.

Item Type:	Article
Divisions :	Surrey research (other units)
Authors :	Koniusz, P, Yan, F and Mikolajczyk, K
Date :	2013
DOI :	10.1016/j.cviu.2012.10.010
Depositing User :	Symplectic Elements
Date Deposited :	28 Mar 2017 13:21
Last Modified :	24 Jan 2020 12:07
URI:	http://epubs.surrey.ac.uk/id/eprint/800379

Actions (login required)

View Item

Downloads