- Updated DP 2010/02 with corrected PDF.
- Temporarily fixed some styling issues with abstracts.
1 parent 56c81f5 commit 8317c253672b85b7a39b9911f8c67eeb27172041
nstanger authored on 4 May 2010
Showing 3 changed files
View
10
Website/dp2010-abstracts-contents.htm
 
<h4>M. Winikoff</h4>
 
<p>In this paper we consider the broader issue of gaining assurance that an agent system will behave appropriately when it is deployed. We ask to what extent this problem is addressed by existing research into formal veri&#64257;cation. We identify a range of issues with existing work which leads us to conclude that, broadly speaking, veri&#64257;cation approaches on their own are too narrowly focussed. We argue that a shift in direction is needed, and outline some possibilities for such a shift in direction.</p>
 
<link rel="Stylesheet" href="DPSstyles.css" type="text/css">
 
<p><a href="papers/dp2010-01.pdf">Download</a> (PDF, 296 KB)</p>
 
<hr>
<p>Novelty detection is an important functionality that has found many applications in information retrieval and processing. In this paper we propose a novel framework that deals with novelty detection for multiple-scene image sets. Working with wildlife image data, the framework starts with image segmentation, followed by feature extraction and classification of the image blocks extracted from image segments. The labelled image blocks are then scanned through to generate a co-occurrence matrix of object labels, representing the semantic context within the scene. The semantic co-occurrence matrices then undergo binarization and principal component analysis for dimension reduction, forming the basis for constructing one-class models for each scene category. An algorithm for outlier detection that employs multiple one-class models is proposed. An advantage of our approach is that it can be used for scene classification and novelty detection at the same time. Our experiments show that the proposed approach algorithm gives favourable performance for the task of detecting novel wildlife scenes, and binarization of the label co-occurrence matrices helps to significantly increase the robustness in dealing with the variation of scene statistics.</p>
 
<p><strong>Keywords:</strong> context, co-occurrence matrix, semantics, novel image, multi-class</p>
 
<p><a href="papers/dp2010-02.pdf">Download</a> (PDF, 3.2 MB)</p>
<p class="note">Corrected version uploaded 2010-05-04.</p>
 
<p><a href="papers/dp2010-02.pdf">Download</a> (PDF, 3.4 MB)</p>
 
<hr>
 
<a name="dp2010-03"></a><h3>2010/03: Norm identification in multi-agent societies</h3>
 
View
2
■■■
Website/dpblurb.htm
<LINK REL="Stylesheet" HREF="DPSstyles.css" TYPE="text/css">
 
<P>These papers represent <EM>work to date at the time of publication</EM> and may not necessarily form the basis for the authors&#8217; final conclusions relating to the topic. It is likely, however, that papers will appear in some form in a journal or in conference proceedings after publication in this Series. Details of final journal/conference publication venues for these papers (where applicable) may be found through the <A HREF="http://www.business.otago.ac.nz/infosci/pubs/">Staff Publications</A> pages.</P>
 
<P>Papers are available in electronic form as indicated. Note that papers published in the Series since 2001 are available in electronic form <EM>only</EM>; hard copies of papers published prior to 2001 may still be available.</P>
 
View
2
■■■
Website/dps2010-include.htm
<p><strong>2010/01:</strong> M. Winikoff. <em>Assurance of agent systems: What role should formal verification play?.</em> <a href="dp2010-abstracts.htm#dp2010-01">Abstract</a> | <a href="papers/dp2010-01.pdf">Download</a> (296 KB)</p>
 
<p><strong>2010/02:</strong> S.-P. Yong and J. D. Deng and M. K. Purvis. <em>Modelling semantic context for novelty detection in wildlife scenes.</em> <a href="dp2010-abstracts.htm#dp2010-02">Abstract</a> | <a href="papers/dp2010-02.pdf">Download</a> (3.2 MB)</p>
<p><strong>2010/02:</strong> S.-P. Yong and J. D. Deng and M. K. Purvis. <em>Modelling semantic context for novelty detection in wildlife scenes.</em> <a href="dp2010-abstracts.htm#dp2010-02">Abstract</a> | <a href="papers/dp2010-02.pdf">Download</a> (3.4 MB) <span class="note">(Corrected version uploaded 2010-05-04.)</span></p>
 
<p><strong>2010/03:</strong> B. T. R. Savarimuthu and S. Cranefield and M. A. Purvis and M. K. Purvis. <em>Norm identification in multi-agent societies.</em> <a href="dp2010-abstracts.htm#dp2010-03">Abstract</a> | <a href="papers/dp2010-03.pdf">Download</a> (562 KB)</p>
 
<p><strong>2010/04:</strong> R. M. Withanawasam and P. A. Whigham and T. Crack and I. M. Premachandra. <em>An empirical investigation of the Maslov limit order market model.</em> <a href="dp2010-abstracts.htm#dp2010-04">Abstract</a> | <a href="papers/dp2010-04.pdf">Download</a> (832 KB)</p>