@techreport{blonde:arxiv2022,
      title={Optimality Inductive Biases and Agnostic Guidelines for Offline Reinforcement Learning}, 
      author={Lionel Blondé and Alexandros Kalousis and Stéphane Marchand-Maillet},
      year={2022},
      eprint={2107.01407},     
	  volume ={abs/2107.01407},
	      journal = { CoRR },
      archivePrefix={arXiv},
	  url = { https://arxiv.org/abs/2107.01407 },
      primaryClass={cs.LG}
	  
}

   @techreport{graziani:arxiv2022,
      title={Learning Interpretable Microscopic Features of Tumor by Multi-task Adversarial CNNs Improves Generalization}, 
      author={Mara Graziani and Sebastian Otalora and Stephane Marchand-Maillet and Henning Muller and Vincent Andrearczyk},
      year={2022},
      eprint={2008.01478},  
	  volume={abs/2008.01478},
	  journal = { CoRR },
 	  url = { https://arxiv.org/abs/2008.01478 },
     archivePrefix={arXiv},
      primaryClass={cs.CV}
}


  @techreport{brangbour:arxiv2022,
  author    = {Etienne Brangbour and
               Pierrick Bruneau and
               Thomas Tamisier and
               Stephane Marchand-Maillet},
  title     = {Cold Start Active Learning Strategies in the Context of Imbalanced
               Classification},
  journal   = {CoRR},
  volume    = {abs/2201.10227},
  year      = {2022},
  url       = {https://arxiv.org/abs/2201.10227}
}

  @techreport{marini:arxiv2022,
  author    = {Niccol{\`{o}} Marini and
               Manfredo Atzori and
               Sebastian Ot{\'{a}}lora and
               St{\'{e}}phane Marchand{-}Maillet and
               Henning M{\"{u}}ller},
  title     = {H{\&}E-adversarial network: a convolutional neural network to
               learn stain-invariant features through Hematoxylin {\&} Eosin
               regression},
  journal   = {CoRR},
  volume    = {abs/2201.06329},
  year      = {2022},
  url       = {https://arxiv.org/abs/2201.06329}
}

  @techreport{messina:arxiv2106.00358,
  author    = {Nicola Messina and
               Giuseppe Amato and
               Fabrizio Falchi and
               Claudio Gennaro and
               Stephane Marchand-Maillet},
  title     = {Towards Efficient Cross-Modal Visual Textual Retrieval using Transformer-Encoder
               Deep Features},
  journal   = {CoRR},
  volume    = {abs/2106.00358},
  year      = {2021},
  url       = {https://arxiv.org/abs/2106.00358}
}



  @techreport{messina:arxiv2008.05231,
	author = {Nicola Messina and Giuseppe Amato and Andrea Esuli and Fabrizio Falchi and Claudio Gennaro and  St{\'{e}}phane Marchand{-}Maillet},
	title = {Fine-grained Visual Textual Alignment for Cross-Modal Retrieval using Transformer Encoders},
	institution = {CoRR abs/2008.05231},
	year = {2020},
       url = {https://arxiv.org/abs/2008.05231}
  }


  @techreport{brangbour:arxiv2012.03731,
	author = {Etienne Brangbour and Pierrick Bruneau and St{\'{e}}phane Marchand{-}Maillet and Renaud Hostache and Marco Chini and Patrick Matgen and Thomas Tamisier},
	title = {Computing flood probabilities using Twitter: application to the Houston urban area during Harvey},
	institution = {CoRR abs/2012.03731},
	year = {2020},
       url = {https://arxiv.org/abs/2012.03731}
  }

  @techreport{ouvrard:arxiv2003.07323,
	author = {Xavier Ouvrard and Jean-Marie Le Goff and Stéphane Marchand-Maillet},
	title = {Tuning Ranking in Co-occurrence Networks with General Biased Exchange-based Diffusion on Hyper-bag-graphs},
	institution = {CoRR abs/2003.07323},
	year = {2020},
       url = {https://arxiv.org/abs/2003.07323}
  }
  

  @techreport{ouvrard:arxiv1905.11695,
	author = {Xavier Ouvrard and Jean-Marie Le Goff and Stéphane Marchand-Maillet},
	title = {The HyperBagGraph DataEdron: An Enriched Browsing Experience of Multimedia Datasets},
	institution = {CoRR abs/1905.11695},
	year = {2019},
       url = {https://arxiv.org/abs/1905.11695}
  }


  @techreport{strasser:arxiv1905.11245,
	author = {Pablo Strasser and Stéphane Armand and Stéphane Marchand-Maillet and Alexandros Kalousis},
	title = {Learning by stochastic serializations},
	institution = {CoRR abs/1905.11245},
	year = {2019},
       url = {https://arxiv.org/abs/1905.11245}
  }


  @techreport{brangbour:arxiv1903.04748,
	author = {Etienne Brangbour and Pierrick Bruneau and Stéphane Marchand-Maillet and Renaud Hostache and Patrick Matgen and Marco Chini and Thomas Tamisier},
	title = {Extracting localized information from a Twitter corpus for flood prevention},
	institution = {CoRR abs/1903.04748},
	year = {2019},
       url = {https://arxiv.org/abs/1903.04748}
  }
  
  @techreport{ouvrard:arxiv1809.00190,
	author = {Xavier Ouvrard and Jean-Marie Le Goff and Stéphane Marchand-Maillet},
	title = {Exchange-Based Diffusion in Hb-Graphs: Highlighting Complex Relationships},
	institution = {CoRR abs/1809.00190},
	year = {2018},
       url = {https://arxiv.org/abs/1809.00190}
  }

  @techreport{ouvrard:arxiv1809.00164,
	author = {Xavier Ouvrard and Jean-Marie Le Goff and Stéphane Marchand-Maillet},
	title = {Hypergraph Modeling and Visualisation of Complex Co-occurence Networks},
	institution = {CoRR abs/1809.00164},
	year = {2018},
       url = {https://arxiv.org/abs/1809.00164}
  }

  @techreport{ouvrard:arxiv1809.00162,
	author = {Xavier Ouvrard and Jean-Marie Le Goff and Stéphane Marchand-Maillet},
	title = {On Adjacency and e-Adjacency in General Hypergraphs: Towards a New e-Adjacency Tensor},
	institution = {CoRR abs/1809.00162},
	year = {2018},
       url = {https://arxiv.org/abs/1809.00162}
  }

  @techreport{gregorova:arxiv1805.06258,
	author = {Magda Gregorová and Alexandros Kalousis and Stéphane Marchand-Maillet},
	title = {Structured nonlinear variable selection},
	institution = {CoRR abs/1805.06258},
	year = {2018},
       url = {https://arxiv.org/abs/1805.06258}
  }

  @techreport{gregorova:arxiv1804.07169,
	author = {Magda Gregorová and Jason Ramapuram and Alexandros Kalousis and Stéphane Marchand-Maillet},
	title = {Large-scale Nonlinear Variable Selection via Kernel Random Features.},
	institution = {CoRR abs/1804.07169},
	year = {2018},
       url = {https://arxiv.org/abs/1804.07169}
  }

  @techreport{ouvrard2017:arxiv1712.08189,
	author = {Xavier Ouvrard and S. Marchand-Maillet},
	title = {Adjacency Matrix and Co-occurrence Tensor of General Hypergraphs: Two Well Separated Notions},
	institution = {CoRR abs/1712.08189},
	year = {2017},
       url = {https://arxiv.org/abs/1712.08189}
  }

   @techreport{nielsen2017:arxiv1701.03916,
	author = {Frank Nielsen and Ke Sun and S. Marchand-Maillet},
	title = {On Hölder projective divergences},
	institution = {CoRR abs/1701.03916},
	year = {2017},
       url = {https://arxiv.org/abs/1701.03916}
  }
  
  @techreport{ouvrard2017:arxiv1707.0015,
	author = {Xavier Ouvrard and Jean-Marie Le Goff and S. Marchand-Maillet},
	title = {Networks of Collaborations: Hypergraph modeling and visualisation},
	institution = {CoRR abs/1707.0015},
	year = {2017},
       url = {https://arxiv.org/abs/1707.0015}
  }
  
  @techreport{gregorova2015:arxiv1706.08811,
	author = {Magda Gregorova and Alexandros Kalousis and S. Marchand-Maillet},
	title = {Forecasting and Granger modelling with non-linear dynamical dependencies},
	institution = {CoRR abs/1706.08811},
	year = {2017},
       url = {https://arxiv.org/abs/1706.08811}
  }

  @techreport{gregorova2015:arxiv1507.01978,
	author = {Magda Gregorova and Alexandros Kalousis and S. Marchand-Maillet},
	title = {Learning Leading Indicators for time series predictions},
	institution = {CoRR abs/1507.01978},
	year = {2015},
       url = {https://arxiv.org/abs/1507.01978}
  }

  @techreport{wang2014:arxiv1405.2798,
	author = {Jun Wang and Ke Sun and Fei Sha and S. Marchand-Maillet and Alexandros Kalousis},
	title = {Two-stage Metric Learning},
	institution = {CoRR abs/1405.2798},
	year = {2014},
       url = {https://arxiv.org/abs/1405.2798}
  }

  @techreport{szekely2009:tr0901,
	author = {Eniko Szekely and Eric Bruno and S. Marchand-Maillet},
	title = {Collection guiding: review of the main strategies fro multimedia collection browsing},
	institution = {Viper group - CVMLab, Dept of Computer Science, University of Geneva},
	year = {2009},
	number = {09.01},
	address = {Route de Drize, 7, CH-1227, Carouge, Switzerland},
       url = {http://viper.unige.ch/documents/pdf/szekely2009-tr0901.pdf}
  }

  @techreport{szekely2008:tr0801,
	author = {Eniko Szekely and Eric Bruno and S. Marchand-Maillet},
	title = {    Unsupervised Dimension Reduction of High-Dimensional Data for Cluster Preservation},
	institution = {Viper group - CVMLab, Dept of Computer Science, University of Geneva},
	year = {2008},
	number = {08.01},
	address = {Route de Drize, 7, CH-1227, Carouge, Switzerland},
       url = {http://viper.unige.ch/documents/pdf/szekely2008-tr0801.pdf}
  }

  @techreport{kosinov2006:tr0601,
	author = {Serhiy Kosinov and S. Marchand-Maillet},
	title = {Dual diffusion model of spreading activation for content-based image
		retrieval},
	institution = {Computer Vision and Multimedia Laboratory, Computing Centre, University
		of Geneva},
	year = {2006},
	number = {06.01},
	address = {Rue G\'en\'eral Dufour, 24, CH-1211, Geneva, Switzerland},
       url = {http://viper.unige.ch/documents/pdf/kosinov2006-tr0601.pdf}
  }

  @techreport{VG:Oct0305,
	author = {Carlo Jelmini and St{\'e}phane Marchand-Maillet},
	title = {The Semantic Web Knowledge Base ({SWKB}), an {OWL} reasoning toolwith retractable inference},
	institution = {Computer Vision and Multimedia Laboratory, Computing Centre, Universityof Geneva},
	year = 2003,
	address = {Rue G\'en\'eral Dufour, 24, CH-1211, Geneva, Switzerland}
  }

  @techreport{VG:Mar0303,
	author = {Tayeb Bouzerda and St{\'e}phane Marchand-Maillet},
	title = {A flexible framework for the development of XML protocols: Applications
		to MRML},
	institution = {Computer Vision and Multimedia Laboratory, Computing Centre, University
		of Geneva},
	year = {2003},
	number = {03.03},
	address = {rue General Dufour, 24, CH-1211, Geneva, Switzerland},
	month = {June},
	vgclass = {report},
	vgproject = {viper}
  }

  @techreport{VG:Dec0307,
	author = {Serhiy Kosinov},
	title = {Visual object recognition using distance-based discriminant analysis},
	institution = {Computer Vision and Multimedia Laboratory, Computing Centre, University
		of Geneva},
	year = {2003},
	number = {03.07},
	address = {Rue G\'en\'eral Dufour, 24, CH-1211, Geneva, Switzerland},
	vgclass = {report},
	vgproject = {viper}
  }

  @techreport{VG:Dec0308,
	author = {St{\'e}phane Marchand-Maillet},
	title = {Collection guiding},
	institution = {Computer Vision and Multimedia Laboratory, Computing Centre, University of Geneva},
	year = 2003,
	address = {Rue G\'en\'eral Dufour, 24, CH-1211, Geneva, Switzerland},
	vgclass = {report},
	vgproject = {viper}
  }

  @techreport{VG:Mar0301,
	author = {St{\'e}phane Marchand-Maillet},
	title = {Meeting Record Modelling for Enhanced Browsing},
	institution = {Computer Vision and Multimedia Laboratory, Computing Centre, University
		of Geneva},
	year = {2003},
	number = {03.01},
	address = {rue General Dufour, 24, CH-1211, Geneva, Switzerland},
	month = {March},
	vgclass = {report},
	vgproject = {viper}
  }

  @techreport{VG:Mar0302,
	author = {St{\'e}phane Marchand-Maillet},
	title = {MRML: Steps towards version 2},
	institution = {Computer Vision and Multimedia Laboratory, Computing Centre, University
		of Geneva},
	year = {2003},
	number = {03.02},
	address = {rue General Dufour, 24, CH-1211, Geneva, Switzerland},
	month = {March},
	vgclass = {report},
	vgproject = {viper}
  }

  @techreport{VG:Oct0304,
	author = {Nicolas Mo{\"e}nne-Loccoz},
	title = {Characterizing activity in video shots based on salient points},
	institution = {Computer Vision and Multimedia Laboratory, Computing Centre, University
		of Geneva},
	year = {2003},
	number = {03.04},
	address = {rue General Dufour, 24, CH-1211, Geneva, Switzerland},
	month = {October},
	vgclass = {report},
	vgproject = {viper}
  }

  @techreport{VG:MMM2001,
	author = {Henning M{\"u}ller and Wolfgang M{\"u}ller and St{\'e}phane Marchand-Maillet
		and David McG. Squire and Thierry Pun},
	title = {Automated benchmarking in content-based image retrieval},
	institution = {University of Geneva},
	year = {2001},
	number = {01.01},
	month = {May},
	url = {http://vision.unige.ch/publications/postscript/2001/MuellerHMuellerWMarchandSquirePun_tr01.pdf},
	abstract = {Benchmarking has always been a crucial problem in content-based image
		retrieval (CBIR). A key issue is the lack of a common access method
		to retrieval systems, such as SQL for relational databases. The Multimedia
		Retrieval Mark-up Language (MRML) solves this problem by standardizing
		access to CBIR systems (CBIRSs). Other difficult problems are also
		shortly addressed, such as obtaining relevance judgments and choosing

		a database for performance comparison. In this article we present
		a fully automated benchmark for CBIRSs based on MRML, which can be
		adapted to any image database and almost any kind of relevance judgment.
		The test evaluates the performance of positive and negative relevance
		feedback, which can be generated automatically from the relevance
		judgments. To illustrate our purpose, a freely available, non-copyright
		image collection is used to evaluate our CBIRS, \emph{Viper}. All
		scripts described here are also freely available for download.},
	url1 = {http://vision.unige.ch/publications/postscript/2001/MuellerHMuellerWMarchandSquirePun_tr01.ps.gz},
	vgclass = {report},
	vgproject = {viper}
  }

  @techreport{VG:MMM2000,
	author = {Wolfgang M{\"u}ller and Henning M{\"u}ller and St{\'e}phane Marchand-Maillet
		and Thierry Pun and David McG. Squire and Zoran Pe\u{c}enovi\'{c}
		and Christoph Giess and Arjen P. de Vries},
	title = {{MRML}: A Communication Protocol for Content-Based Image Retrieval},
	institution = {Computer Vision Group, Computing Centre, University of Geneva},
	year = {2000},
	number = {00.02},
	address = {rue G\'en\'eral Dufour, 24, CH-1211 Gen\`eve, Switzerland},
	month = {mar},
	url = {http://vision.unige.ch/publications/postscript/2000/VGTR00.02_MuellerWMuellerHMarchandPunSquireGiessVries.ps.gz},
	abstract = {In this paper we introduce and describe the Multimedia Retrieval Markup
		Language (MRML). This XML-based markup language is the basis for
		an open communication protocol for content-based image retrieval
		systems (CBIRSs). MRML was initially designed as a means of separating
		CBIR engines from their user interfaces. It is, however, also extensible
		as the basis for standardized performance evaluation procedures.
		Such a tool is essential for the formulation and implementation of
		common benchmarks for CBIR. A common protocol can also bring new
		dynamics to the CBIR field---it makes the development of new systems
		faster and more efficient, and opens the door of the CBIR research
		field to other disciplines such as Human-Computer Interaction. The
		MRML specifications, as well as the first MRML-compliant applications,
		are freely available and are introduced in this paper.},
	url1 = {http://vision.unige.ch/publications/postscript/2000/VGTR00.02_MuellerWMuellerHMarchandPunSquireGiessVries.pdf},
	vgclass = {report},
	vgproject = {viper}
  }

  @techreport{VG:MMS2000a,
	author = {Henning M{\"u}ller and Wolfgang M{\"u}ller and David McG. Squire
		and Zoran Pe\u{c}enovi\'{c} and St{\'e}phane Marchand-Maillet and
		Thierry Pun},
	title = {An Open Framework for Distributed Multimedia Retrieval},
	institution = {Computer Vision Group, Computing Centre, University of Geneva},
	year = {2000},
	number = {00.03},
	address = {rue G\'en\'eral Dufour, 24, CH-1211 Gen\`eve, Switzerland},
	month = {mar},
	url = {http://vision.unige.ch/publications/postscript/2000/VGTR00.03_MuellerHMuellerWSquirePecenovicMarchandPun.ps.gz},
	abstract = {This article describes a framework for distributed multimedia retrieval
		which permits the connection of compliant user interfaces with a
		variety of multimedia retrieval engines via an open communication
		protocol, MRML (Multi Media Retrieval Markup Language). It allows
		the choice of image collection, feature set and query algorithm during
		run--time, permitting multiple users to query a system adapted to
		their needs, using the query paradigm adapted to their problem such
		as query by example (QBE), browsing queries, or query by annotation.
		User interaction is implemented over several levels and in diverse
		ways. Relevance feedback is implemented using positive and negative
		example images that can be used for a best--match QBE query. In contrast,
		browsing methods try to ap proach the searched image by giving overviews
		of the entire collection and by successive refinements. In addition
		to these query methods, Long term off line learning is implemented.
		It allows feature preferences per user, user domain or over all users
		to be learned automatically. We present the Viper multimedia retrieval
		system as the core of the framework and an example of an MRML-compliant
		search engine. Viper uses techniques adapted from traditional information
		retrieval (IR) to retrieve multimedia documents, thus benefiting
		from the many years of IR research. As a result, textual and visual
		features are treated in the same way, facilitating true multimedia
		retrieval. The MRML protocol also allows other applications to make
		use of the search engi nes. This can for example be used for the
		design of a benchmark test suite, querying several search engines
		in the same way and comparing the results. This is motivated by the
		fact that the content--based image retrieval community really lacks
		such a benchmark as it already exists in text retrieval.},
	url1 = {http://vision.unige.ch/publications/postscript/2000/VGTR00.03_MuellerHMuellerWSquirePecenovicMarchandPun.pdf},
	vgclass = {report},
	vgproject = {viper}
  }

  @techreport{VG:MMS2000b,
	author = {Henning M{\"u}ller and Wolfgang M{\"u}ller and David McG. Squire
		and St{\'e}phane Marchand-Maillet and Thierry Pun},
	title = {Long-Term Learning from User Behavior in Content-Based Image Retrieval},
	institution = {Computer Vision Group, Computing Centre, University of Geneva},
	year = {2000},
	number = {00.04},
	address = {rue G\'en\'eral Dufour, 24, CH-1211 Gen\`eve, Switzerland},
	month = {mar},
	url = {http://vision.unige.ch/publications/postscript/2000/VGTR00.04_MuellerHMuellerWSquireMarchandPun.ps.gz},
	abstract = {This article describes a simple algorithm for obtaining knowledge
		about the importance of features from analyzing user log files of
		a content-based image retrieval system (CBIRS). The user log files
		of the usage of the Viper web demonstration system are analyzed over
		a period of four months. In this time about 3500 accesses to the
		system were made with 800 multiple image queries. The analysis only
		takes into account multiple image queries of the system with positive
		or negative input images, because only these queries contain enough
		information for the method described in the paper. Features frequently
		present in images marked together positively in the same query step
		get a higher weighting whereas features present in an image marked
		positively and another image marked negatively in the same step get
		a lower weighting. The Viper system offers a very large number of
		simple features which allows the creation of feature weightings with
		high values for important and low values for less important features.
		These weightings for features can of course differ for several collections
		and as well for several users. The results are evaluated using the
		relevance judgments of real users and compared to the system without
		the long-term learning.},
	url1 = {http://vision.unige.ch/publications/postscript/2000/VGTR00.04_MuellerHMuellerWSquireMarchandPun.pdf},
	vgclass = {report},
	vgproject = {viper}
  }

  @techreport{VG:MMS2000,
	author = {Henning M{\"u}ller and Wolfgang M{\"u}ller and David McG. Squire
		and St{\'e}phane Marchand-Maillet and Thierry Pun},
	title = {Strategies for positive and negative relevance feedback in image
		retrieval},
	institution = {Computer Vision Group, Computing Centre, University of Geneva},
	year = {2000},
	number = {00.01},
	address = {rue G\'en\'eral Dufour, 24, CH-1211 Gen\`eve, Switzerland},
	month = {jan},
	url = {http://vision.unige.ch/publications/postscript/2000/VGTR00.01_HMuellerWMuellerSquireMarchandPun.ps},
	abstract = {Relevance feedback has been shown to be a very effective tool for
		enhancing retrieval results in text retrieval. In content-based image
		retrieval it is more and more frequently used and very good results
		have been obtained. However, too much negative feedback may destroy
		a query as good features get negative weightings. This paper compares
		a variety of strategies for positive and negative feedback. The performance
		evaluation of feedback algorithms is a hard problem. To solve this,
		we obtain judgments from several users and employ an automated feedback
		scheme. We can then evaluate different techniques using the same
		judgments. Using automated feedback, the ability of a system to adapt
		to the user's needs can be measured very effectively. Our study highlights
		the utility of negative feedback, especially over several feedback
		steps.},
	url1 = {http://vision.unige.ch/publications/postscript/2000/VGTR00.01_HMuellerWMuellerSquireMarchandPun.pdf},
	vgclass = {report},
	vgproject = {viper}
  }

  @techreport{VG:Mar2000,
	author = {St{\'e}phane Marchand-Maillet},
	title = {Content-Based Video Retrieval: An Overview},
	institution = {CUI - University of Geneva},
	year = {2000},
	number = {00.06},
	address = {Geneva, Switzerland},
	url = {http://vision.unige.ch/publications/postscript/2000/VGTR00.06_Marchand.pdf},
	abstract = {Content-based Image Retrieval systems (CBIRS) start flourishing on
		the Web. Their performances are continuously improving and their
		base principles span a wide range of diversity. Content-based Video
		Retrieval systems (CBVRS) are less common and seem at a first glance
		to be a natural extension of CBIRS. In this document, we summarise
		advances made in the development of CBVRS and analyse their relationship
		to CBIRS. While doing so, we show that CBVRS are actually not so
		obvious extensions of CBIRS. (40 References)},
	url1 = {http://vision.unige.ch/publications/postscript/2000/VGTR00.06_Marchand.pdf},
	url2 = {http://viper.unige.ch/~marchand/CBVR/},
	vgclass = {report},
	vgproject = {viper}
  }

  @techreport{VG:MMS1999,
	author = {Henning M{\"u}ller and Wolfgang M{\"u}ller and David McG. Squire
		and Thierry Pun},
	title = {Performance Evaluation in Content-Based Image Retrieval: Overview
		and Proposals},
	institution = {Computer Vision Group, Computing Centre, University of Geneva},
	year = {1999},
	number = {99.05},
	address = {rue G\'en\'eral Dufour, 24, CH-1211 Gen\`eve, Switzerland},
	month = {dec},
	url = {http://vision.unige.ch/publications/postscript/99/VGTR99.05_HMuellerWMuellerSquirePun.ps.gz},
	abstract = {Evaluation of retrieval performance is a crucial problem in content-based
		image retrieval (CBIR). Many different methods for measuring the
		performance of a system have been created and used by researchers.
		This article discusses the advantages and shortcomings of the performance
		measures currently used. Problems such as a common image database
		for performance comparisons and a means of getting relevance judgments
		(or ground truth) for queries are explained. The relationship between
		CBIR and information retrieval (IR) is made clear, since IR researchers
		have decades of experience with the evaluation problem. Many of their
		solutions can be used for CBIR, despite the differences between the
		fields. Several methods used in text retrieval are explained. Proposals
		for performance measures and means of developing a standard test
		suite for CBIR, similar to that used in IR at the annual Text REtrieval
		Conference (TREC), are presented.},
	url1 = {http://vision.unige.ch/publications/postscript/99/VGTR99.05_HMuellerWMuellerSquirePun.pdf},
	vgclass = {report},
	vgproject = {viper}
  }

  @techreport{VG:MSM1999,
	author = {Henning M{\"u}ller and David McG. Squire and Wolfgang M{\"u}ller
		and Thierry Pun},
	title = {Efficient access methods for content-based image retrieval with inverted
		files},
	institution = {Computer Vision Group, Computing Centre, University of Geneva},
	year = {1999},
	number = {99.02},
	address = {rue G\'en\'eral Dufour, 24, CH-1211 Gen\`eve, Switzerland},
	month = {July},
	url = {http://vision.unige.ch/publications/postscript/99/VGTR99.02_MuellerSquireMuellerPun.ps.gz},
	abstract = {As human factor studies over the last thirty years have shown, response
		time is a very important factor for the usability of an interactive
		system, especially on the world wide web. In particular, response
		times of under one second are often specified as a usability requirement
		\cite{Nie97}. This paper compares several methods for improving the
		evaluation time in a content-based image retrieval system (CBIRS)
		which uses inverted file technology. The use of the inverted file
		technology facilitates search pruning in a variety of ways, as is
		shown in this paper. For large databases ($> 2000$ images) and a
		high number of possible features ($> 80000$), efficient and fast
		access is necessary to allow interactive querying and browsing. Parallel
		access to the inverted file can reduce the response time. This parallel
		access is very easy to implement with little communication overhead,
		and thus scales well. Other search pruning methods, similar to methods
		used in information retrieval, can also reduce the response time
		significantly without reducing the performance of the system. The
		performance of the system is evaluated using precision vs. recall
		graphs, which are an established evaluation method in information
		retrieval. A user survey was carried out in order to obtain relevance
		judgments for the queries reported in this work.},
	keywords = {inverted file, content-based image retrieval, efficient access, search
		pruning, speed evaluation},
	url1 = {http://vision.unige.ch/publications/postscript/99/VGTR99.02_MuellerSquireMuellerPun.pdf},
	vgclass = {report},
	vgproject = {viper}
  }

  @techreport{VG:MPV1999,
	author = {Wolfgang M{\"u}ller and Zoran Pe\u{c}enovi\'{c} and Arjen P. de Vries
		and David McG. Squire and Henning M{\"u}ller and Thierry Pun},
	title = {{MRML}: Towards an extensible standard for multimedia querying and
		benchmarking (Draft Proposal)},
	institution = {Computer Vision Group, Computing Centre, University of Geneva},
	year = {1999},
	number = {99.04},
	address = {rue G\'en\'eral Dufour, 24, CH-1211 Gen\`eve, Switzerland},
	month = {October},
	url = {http://vision.unige.ch/publications/postscript/99/VGTR99.04_WMuellerPecenovicdeVriesSquireHMuellerPun.ps.gz},
	url1 = {http://vision.unige.ch/publications/postscript/99/VGTR99.04_WMuellerPecenovicdeVriesSquireHMuellerPun.pdf},
	vgclass = {report},
	vgproject = {viper}
  }

  @techreport{VG:MSM1999a,
	author = {Wolfgang M{\"u}ller and David McG. Squire and Henning M{\"u}ller
		and Thierry Pun},
	title = {Hunting moving targets: an extension to {B}ayesian methods in multimedia
		databases},
	institution = {Computer Vision Group, Computing Centre, University of Geneva},
	year = {1999},
	number = {99.03},
	address = {rue G\'en\'eral Dufour, 24, CH-1211 Gen\`eve, Switzerland},
	month = {July},
	url = {http://vision.unige.ch/publications/postscript/99/VGTR99.03_MuellerSquireMuellerPun.ps.gz},
	abstract = {It has been widely recognised that the difference between the level
		of abstraction of the formulation of a query (by example) and that
		of the desired result (usually an image with certain semantics) calls
		for the use of learning methods that try to bridge this gap. Cox
		\emph{et al.}~have proposed a Bayesian method to learn the user's
		preferences during each query. Cox \emph{et al.}\'s system, \texttt{PicHunter},
		is designed for optimal performance when the user is searching for
		a fixed target image. The performance of the system was evaluated
		using target testing, which ranks systems according to the number
		of interaction steps required to find the target, leading to simple,
		easily reproducible experiments. There are some aspects of image
		retrieval, however, which are not captured by this measure. In particular,
		the possibility of query drift (i.e.~a moving target) is completely
		ignored. The algorithm proposed by Cox \emph{et al.}~does not cope
		well with a change of target at a late query stage, because it is
		assumed that user feedback is noisy, but consistent. In the case
		of a moving target, however, the feedback is noisy \emph{and} inconsistent
		with earlier feedback. In this paper we propose an enhanced Bayesian
		scheme which selectively forgets inconsistent user feedback, thus
		enabling both the program and the user to ``change their minds''.
		The effectiveness of this scheme is demonstrated in moving target
		tests on a database of heterogeneous real-world images.},
	keywords = {relevance feedback, query drift, target testing, Bayesian methods,
		user modelling},
	url1 = {http://vision.unige.ch/publications/postscript/99/VGTR99.03_MuellerSquireMuellerPun.pdf},
	vgclass = {report},
	vgproject = {viper}
  }

  @techreport{VG:SMM1999b,
	author = {David McG. Squire and Henning M{\"u}ller and Wolfgang M{\"u}ller},
	title = {Improving Response Time by Search Pruning in a Content-Based Image
		Retrieval System, Using Inverted File Techniques},
	institution = {Computer Vision Group, Computing Centre, University of Geneva},
	year = {1999},
	number = {99.01},
	address = {rue G\'en\'eral Dufour, 24, CH-1211 Gen\`eve, Switzerland},
	month = {February},
	url = {http://vision.unige.ch/publications/postscript/99/VGTR99.01_SquireMuellerMueller.ps.gz},
	abstract = {This paper describes several methods for improving query evaluation
		speed in a content-based image retrieval system (CBIRS). Response
		time is an extremely important factor in determining the usefulness
		of any interactive system, as has been demonstrated by human factors
		studies over the past thirty years. In particular, response times
		of less than one second are often specified as a usability requirement.
		It is shown that the use of inverted files facilitates the reduction
		of query evaluation time without significantly reducing the accuracy
		of the response. The performance of the system is evaluated using
		precision \vs recall graphs, which are an established evaluation
		method in information retrieval (IR), and are beginning to be used
		by CBIR researchers.},
	keywords = {content-based image retrieval, search pruning, inverted file, response
		time},
	url1 = {http://vision.unige.ch/publications/postscript/99/VGTR99.01_SquireMuellerMueller.pdf},
	vgclass = {report},
	vgproject = {viper}
  }

  @techreport{VG:Squ1998,
	author = {David McG. Squire},
	title = {Generalization performance of factor analysis techniques used for
		image database organization},
	institution = {Computer Vision Group, Computing Centre, University of Geneva},
	year = {1998},
	number = {98.01},
	address = {rue G\'en\'eral Dufour, 24, CH-1211 Gen\`eve, Switzerland},
	month = {January},
	url = {http://vision.unige.ch/publications/postscript/98/VGTR98.01_Squire.ps.gz},
	abstract = {The goal of this paper is to evaluate the generalization performance
		of a variety of factor analysis techniques in an image database environment.
		Factor analysis techniques, such as Principal Components Analysis,
		have been proposed as means of reducing the dimensionality of the
		data stored in image retrieval systems. These techniques compute
		a transformation which is applied to vectors of image features to
		produce vectors of lower dimensionality which still characterize
		the original data well. Computing such transformations for very large
		numbers of images is computationally expensive, especially if this
		calculation must be repeated each time new images are added to the
		database. It is to be hoped, therefore, that a transformation computed
		using a subset of all possible images will perform well when applied
		to images not used in its derivation. To evaluate this generalization
		ability, we measure the agreement between partitionings of image
		sets computed using such transformations with those produced by human
		subjects.},
	url1 = {http://vision.unige.ch/publications/postscript/98/VGTR98.01_Squire.pdf},
	vgclass = {report},
	vgproject = {viper}
  }

  @techreport{VG:Squ1998a,
	author = {David McG. Squire},
	title = {Learning a similarity-based distance measure for image database organization
		from human partitionings of an image set},
	institution = {Computer Vision Group, Computing Centre, University of Geneva},
	year = {1998},
	number = {98.03},
	address = {rue G\'en\'eral Dufour, 24, CH-1211 Gen\`eve, Switzerland},
	month = {April},
	url = {http://vision.unige.ch/publications/postscript/98/VGTR98.03_Squire.ps.gz},
	abstract = {In this paper we employ human judgments of image similarity to improve
		the organization of an image database. We first derive a statistic,
		$\kappa_B$ which measures the agreement between two partitionings
		of an image set. $\kappa_B$ is used to assess agreement both amongst
		and between human and machine partitionings. This provides a rigorous
		means of choosing between competing image database organization systems,
		and of assessing the performance of such systems with respect to
		human judgments. Human partitionings of an image set are used to
		define an similarity value based on the frequency with which images
		are judged to be similar. When this measure is used to partition
		an image set using a clustering technique, the resultant partitioning
		agrees better with human partitionings than any of the feature-space-based
		techniques investigated. Finally, we investigate the use multilayer
		perceptrons and a \emph{Distance Learning Network} to learn a mapping
		from feature space to this perceptual similarity space. The Distance
		Learning Network is shown to learn a mapping which results in partitionings
		in excellent agreement with those produced by human subjects.},
	url1 = {http://vision.unige.ch/publications/postscript/98/VGTR98.03_Squire.pdf},
	vgclass = {report},
	vgproject = {viper}
  }

  @techreport{VG:SMM1998a,
	author = {David McG. Squire and Wolfgang M{\"u}ller and Henning M{\"u}ller},
	title = {Relevance feedback and term weighting schemes for content-based image
		retrieval},
	institution = {Computer Vision Group, Computing Centre, University of Geneva},
	year = {1998},
	number = {98.05},
	address = {rue G\'en\'eral Dufour, 24, CH-1211 Gen\`eve, Switzerland},
	month = {December},
	url = {http://vision.unige.ch/publications/postscript/98/VGTR98.05_SquireMuellerMueller.ps.gz},
	abstract = {This paper describes the application of techniques derived from text
		retrieval research to the content-based querying of image databases.
		Specifically, the use of inverted files, frequency-based weights
		and relevance feedback are investigated. The use of inverted files
		allows very large numbers ($\geq \mathcal{O}(104)$) of \emph{possible}
		features to be used. since search is limited to the subspace spanned
		by the features present in the query image(s). A variety of weighting
		schemes used in text retrieval are employed, yielding different results.
		We suggest possibles modifications for their use with image databases.
		The use of relevance feedback was shown to improve the query results
		significantly, as measured by precision and recall, for all users.},
	url1 = {http://vision.unige.ch/publications/postscript/98/VGTR98.05_SquireMuellerMueller.pdf},
	vgclass = {report},
	vgproject = {viper}
  }

  @techreport{VG:SMM1998,
	author = {David McG. Squire and Wolfgang M{\"u}ller and Henning M{\"u}ller
		and Jilali Raki},
	title = {Content-based query of image databases, inspirations from text retrieval:
		inverted files, frequency-based weights and relevance feedback},
	institution = {Computer Vision Group, Computing Centre, University of Geneva},
	year = {1998},
	number = {98.04},
	address = {rue G\'en\'eral Dufour, 24, CH-1211 Gen\`eve, Switzerland},
	month = {November},
	url = {http://vision.unige.ch/publications/postscript/98/VGTR98.04_SquireMuellerMuellerRaki.ps.gz},
	abstract = {In this paper we report the application of techniques inspired by
		text retrieval research to the content-based query of image databases.
		In particular, we show how the use of an inverted file data structure
		permits the use of a feature space of $\mathcal{O}(104)$ dimensions,
		by restricting search to the subspace spanned by the features present
		in the query. A suitably sparse set of colour and texture features
		is proposed. A scheme based on the frequency of occurrence of features
		in both individual images and in the whole collection provides a
		means of weighting possibly incommensurate features in a compatible
		manner, and naturally extends to incorporate relevance feedback queries.
		The use of relevance feedback is shown consistently to improve system
		performance, as measured by precision and recall.},
	url1 = {http://vision.unige.ch/publications/postscript/98/VGTR98.04_SquireMuellerMuellerRaki.pdf},
	vgclass = {report},
	vgproject = {viper}
  }

  @techreport{VG:SqC1997,
	author = {David McG. Squire and Terry M. Caelli},
	title = {Invariance Signatures: Characterizing contours by their departures
		from invariance},
	institution = {Computer Vision Group, Computing Centre, University of Geneva},
	year = {1997},
	number = {97.04},
	address = {rue G\'en\'eral Dufour, 24, CH-1211 Gen\`eve, Switzerland},
	month = {April},
	url = {http://vision.unige.ch/publications/postscript/97/VGTR97.04_SquireCaelli.ps.gz},
	url1 = {http://vision.unige.ch/publications/postscript/97/VGTR97.04_SquireCaelli.pdf},
	vgclass = {report},
	vgproject = {unspecified}
  }

  @techreport{VG:SqP1997a,
	author = {David McG. Squire and Thierry Pun},
	title = {Assessing Agreement Between Human and Machine Clusterings of Image
		Databases},
	institution = {Computer Vision Group, Computing Centre, University of Geneva},
	year = {1997},
	number = {97.03},
	address = {rue G\'en\'eral Dufour, 24, CH-1211 Gen\`eve, Switzerland},
	month = {April},
	url = {http://vision.unige.ch/publications/postscript/97/VGTR97.03_SquirePun.ps.gz},
	abstract = {There is currently much interest in the organization and \emph{content-based}
		querying image databases. The usual hypothesis is that image similarity
		can be characterized by low-level features, without further abstraction.
		This assumes that agreement between machine and human measures of
		similarity is sufficient for the database to be useful. To assess
		this assumption, we develop measures of the agreement between partitionings
		of an image set, showing that chance agreements \emph{must} be considered.
		These measures are used to assess the agreement between human subjects
		and several machine clustering techniques on an image set. The results
		can be used to select and refine distance measures for querying and
		organizing image databases.},
	url1 = {http://vision.unige.ch/publications/postscript/97/VGTR97.03_SquirePun.pdf},
	vgclass = {report},
	vgproject = {viper}
  }

  @techreport{VG:SqP1997b,
	author = {David McG. Squire and Thierry Pun},
	title = {Using human partitionings of an image set to learn a similarity-based
		distance measure},
	institution = {Computer Vision Group, Computing Centre, University of Geneva},
	year = {1997},
	number = {97.06},
	address = {rue G\'en\'eral Dufour, 24, CH-1211 Gen\`eve, Switzerland},
	month = {November},
	url = {http://vision.unige.ch/publications/postscript/97/VGTR97.06_SquirePun.ps.gz},
	abstract = {In this paper our goal is to employ human judgments of image similarity
		to improve the organization of an image database for content-based
		retrieval. We first derive a statistic, $\kappa_B$ for measuring
		the agreement between two partitionings of an image set into unlabeled
		subsets. This measure can be used both to measure the degree of agreement
		between pairs of human subjects, and also between human and machine
		partitionings of an image set. This provides a rigorous means of
		selecting between competing image database organization systems,
		and assessing how close the performance of such systems is to that
		which might be expected from a database organization done by hand.
		We then use the results of experiments in which human subjects are
		asked to partition a set of images into unlabeled subsets to define
		a similarity measure for pairs of images based on the frequency with
		which they were judged to be similar. We show that, when this measure
		is used to partition an image set using a clustering technique, the
		resultant clustering agrees better with those produced by human subjects
		than any of the feature space-based techniques investigated. Finally,
		we investigate the use of machine learning techniques to discover
		a mapping from a numerical feature space to this perceptual similarity
		space. Such a mapping would allow the ground truth knowledge abstracted
		from the human judgments to be generalized to unseen images.},
	url1 = {http://vision.unige.ch/publications/postscript/97/VGTR97.06_SquirePun.pdf},
	vgclass = {report},
	vgproject = {viper}
  }

  @techreport{VG:Gar1995,
	author = {Catherine De Garrini},
	title = {Exploratory statistics for automated structuration of large image
		databases},
	institution = {AI and Vision Group, Computing Centre, University of Geneva},
	year = {1995},
	number = {95.01},
	address = {rue G\'en\'eral Dufour, 24, CH-1211 Gen\`eve, Switzerland},
	month = {September},
	vgclass = {report},
	vgproject = {cbir}
  }

  @techreport{VG:Gar1994,
	author = {Catherine De Garrini},
	title = {D\'etermination de la translation, de la rotation et du facteur d'echelle
		d'un objet dans differents contextes},
	institution = {AI and Vision Group, Computing Centre, University of Geneva},
	year = {1994},
	number = {94.08},
	address = {rue G\'en\'eral Dufour, 24, CH-1211 Gen\`eve, Switzerland},
	month = {October},
	url = {ftp://cui.unige.ch/PUBLIC/vision/papers/degarrin/94.10.report.ps.Z},
	vgclass = {report},
	vgproject = {unspecified}
  }