Matching entries: 0
settings...
Author Title Year Journal/Proceedings Reftype DOI/URL
Ferreira, C.D., Calumby, R.T., do C. Araujo, I.B.A., Dourado, Í.C., Muñoz, J.A.V., Penatti, O.A.B., Li, L.T., Almeida, J. and da Silva Torres, R. Recod MediaEval 2016: Diverse Social Images Retrieval 2016
Vol. 1739Working Notes Proceedings of the MediaEval 2016 Workshop 
inproceedings URL 
Abstract: This paper presents the RECOD team experience in the Retrieving Diverse Social Images Task at MediaEval 2016. The teams were required to develop a diversification approach for social photo retrieval. Our proposal is based on re-ranking, rank aggregation, and diversity promotion, allowing employment of textual and visual information apart or fused.

BibTeX:
@inproceedings{Ferreira2016MediaEval,
  author = {Cristiano D. Ferreira and Rodrigo Tripodi Calumby and Iago B. A. do C. Araujo and Ícaro C. Dourado and Javier A. V. Muñoz and Otávio Augusto Bizetto Penatti and Lin Tzy Li and Jurandy Almeida and Ricardo da Silva Torres},
  title = {Recod  MediaEval 2016: Diverse Social Images Retrieval},
  booktitle = {Working Notes Proceedings of the MediaEval 2016 Workshop},
  publisher = {CEUR-WS.org},
  year = {2016},
  volume = {1739},
  url = {http://ceur-ws.org/Vol-1739/MediaEval_2016_paper_21.pdf}
}
      


Muñoz, J.A.V., Li, L.T., Dourado, Í.C., Nogueira, K., Fadel, S.G., Penatti, O.A.B., Almeida, J., Pereira, L.A.M., Calumby, R.T., dos Santos, J.A. and da Silva Torres, R. RECOD Placing Task of MediaEval 2016: A Ranking Fusion Approach for Geographic-Location Prediction of Multimedia Objects 2016
Vol. 1739Working Notes Proceedings of the MediaEval 2016 Workshop 
inproceedings URL 
Abstract: We describe the approach proposed by the RECOD team for the estimation-based sub-task of Placing Task at MediaEval 2016. Our approach uses genetic programming (GP) to combine ranked lists de ned in terms of textual and visual descriptors to automatically assign geographic locations to images and videos.

BibTeX:
@inproceedings{Munoz2016MediaEval,
  author = {Javier A. V. Muñoz and Lin Tzy Li and Ícaro C. Dourado and Keiller Nogueira and Samuel G. Fadel and Otávio Augusto Bizetto Penatti and Jurandy Almeida and Luis A. M. Pereira and Rodrigo Tripodi Calumby and Jefersson A. dos Santos and Ricardo da Silva Torres},
  title = {RECOD  Placing Task of MediaEval 2016: A Ranking Fusion Approach for Geographic-Location Prediction of Multimedia Objects},
  booktitle = {Working Notes Proceedings of the MediaEval 2016 Workshop},
  publisher = {CEUR-WS.org},
  year = {2016},
  volume = {1739},
  url = {http://ceur-ws.org/Vol-1739/MediaEval_2016_paper_22.pdf}
}
      


Pedronette, D.C.G. and da Silva Torres, R. Rank Diffusion for Context-Based Image Retrieval 2016 Proceedings of the 2016 ACM on International Conference on Multimedia Retrieval, ICMR, pp. 321-325  inproceedings DOI URL 
Abstract: This paper presents an efficient diffusion-based re-ranking approach. The proposed method propagates contextual information defined in terms of top-ranked objects of ranked lists in a diffusion process. That makes the method suitable for large scale real-world collections. Experiments were conducted considering public image collections, several descriptors, and comparisons with state-of-the-art methods. Experimental results demonstrate that the proposed method provides high effectiveness gains with low computational costs.

BibTeX:
@inproceedings{Pedronette2016ICMR,
  author = {Daniel Carlos Guimarães Pedronette and Ricardo da Silva Torres},
  title = {Rank Diffusion for Context-Based Image Retrieval},
  booktitle = {Proceedings of the 2016 ACM on International Conference on Multimedia Retrieval, ICMR},
  year = {2016},
  pages = {321--325},
  url = {http://doi.acm.org/10.1145/2911996.2912060},
  doi = {http://doi.org/10.1145/2911996.2912060}
}
      


Calumby, R.T., do C. Araujo, I.B.A., Santana, V.P., Muñoz, J.A.V., Penatti, O.A.B., Li, L.T., Almeida, J., Chiachia, G., Gonçalves, M.A. and da Silva Torres, R. Recod MediaEval 2015: Diverse Social Images Retrieval 2015 Working Notes Proceedings of the MediaEval 2015 Workshop  inproceedings  
Abstract: This paper presents the RECOD team experience in the Retrieving Diverse Social Images Task at MediaEval 2015. The teams were required to develop a diversification approach for social photo retrieval. Our proposal is based on irrelevant image filtering, reranking, rank aggregation, and diversity promotion. We proposed a multimodal approach and exploited image metadata and user credibility information.

BibTeX:
@inproceedings{Calumby2015MediaEval,
  author = {Rodrigo Tripodi Calumby and Iago B. A. do C. Araujo and Vin\icius Pereira Santana and Javier A. V. Muñoz and Otávio Augusto Bizetto Penatti and Lin Tzy Li and Jurandy Almeida and Giovani Chiachia and Marcos André Gonçalves and Ricardo da Silva Torres},
  title = {Recod  MediaEval 2015: Diverse Social Images Retrieval},
  booktitle = {Working Notes Proceedings of the MediaEval 2015 Workshop},
  year = {2015}
}
      


Li, L.T., Muñoz, J.A.V., Almeida, J., Calumby, R.T., Penatti, O.A.B., Dourado, Í.C., Nogueira, K., Junior, P.R.M., Pereira, L.A.M., Pedronette, D.C.G., dos Santos, J.A., Gonçalves, M.A. and da Silva Torres, R. RECOD Placing Task of MediaEval 2015 2015 Working Notes Proceedings of the MediaEval 2015 Workshop  inproceedings  
Abstract: In this work, we describe the approach proposed by the RECOD team for the Placing Task, Locale-based sub-task, at MediaEval 2015. Our approach is based on the use of as much evidence as possible (textual, visual, and/or audio descriptors) to automatically assign geographical locations to images and videos.

BibTeX:
@inproceedings{Li2015MediaEval,
  author = {Lin Tzy Li and Javier A. V. Muñoz and Jurandy Almeida and Rodrigo Tripodi Calumby and Otávio Augusto Bizetto Penatti and Ícaro C. Dourado and Keiller Nogueira and Pedro Ribeiro Mendes Junior and Luis A. M. Pereira and Daniel Carlos Guimarães Pedronette and Jefersson Alex dos Santos and Marcos André Gonçalves and Ricardo da Silva Torres},
  title = {RECOD  Placing Task of MediaEval 2015},
  booktitle = {Working Notes Proceedings of the MediaEval 2015 Workshop},
  year = {2015}
}
      


Calumby, R.T., Santana, V.P., Cordeiro, F.S., Penatti, O.A.B., Li, L.T., Chiachia, G. and da Silva Torres, R. Recod MediaEval 2014: Diverse Social Images Retrieval 2014 Working Notes Proceedings of the MediaEval 2014 Workshop, Barcelona, Catalunya, Spain, October 16-17, 2014.  inproceedings URL 
Abstract: This paper presents the results of the first participation of our multi-institutional team in the Retrieving Diverse Social Images Task at MediaEval 2014. In this task we were required to develop a summarization and diversification approach for social photo retrieval. Our approach is based on irrelevant image filtering, image re-ranking, and diversity promotion by clustering. We have used visual and textual features, including image metadata and user credibility information.

BibTeX:
@inproceedings{Calumby2014MediaEval,
  author = {Rodrigo Tripodi Calumby and Vin\icius Pereira Santana and Felipe Souza Cordeiro and Otávio Augusto Bizetto Penatti and Lin Tzy Li and Giovani Chiachia and Ricardo da Silva Torres},
  title = {Recod  MediaEval 2014: Diverse Social Images Retrieval},
  booktitle = {Working Notes Proceedings of the MediaEval 2014 Workshop, Barcelona, Catalunya, Spain, October 16-17, 2014.},
  year = {2014},
  url = {http://ceur-ws.org/Vol-1263/mediaeval2014_submission_38.pdf}
}
      


Kozievitch, N., Hara, C.S., Nande, J. and da Ricardo, S Exploring Data Fusion under the Image Retrieval Domain 2014 16th International Conference on Enterprise Information Systems  inproceedings  
Abstract: Advanced services in data compression, data storage, and data transmission have been developed and are widely used to address the required capabilities of an assortment of systems across diverse application domains. In order to reuse, integrate, unify, manage, and support heterogeneous resources, a number of works and concepts have emerged with the aim of facilitating aggregation of content and helping system developers. In particular, images, along with existing Content-Based Image Retrieval services, have the potential to play a key role in information systems, due to the large availability of images and the need to integrate them with existing collections, metadata, and available image manipulation softwares and applications. In this work, we explore a data fusion approach for solving data value conflicts in the context of image retrieval domain. In particular, we target the process of solving value conflicts resulted from different features integrating the data resulted from the Content-Based Image Retrieval process, along with the image metadata, provided from a number of sources and applications. Our approach reduces the need of human intervention for keeping a clean and integrated view of an image repository when new data sources are added to an image management system.

BibTeX:
@inproceedings{Kozievitch2014ICEIS,
  author = {Nádia Kozievitch and Carmen Satie Hara and Jaqueline Nande and Ricardo da S},
  title = {Exploring Data Fusion under the Image Retrieval Domain},
  booktitle = {16th International Conference on Enterprise Information Systems},
  year = {2014}
}
      


Li, L.T., Penatti, O.A.B., Almeida, J., Chiachia, G., Calumby, R.T., Mendes-Junior, P.R., Pedronette, D.C.G. and da Silva Torres, R. Multimedia Geocoding: The RECOD 2014 Approach 2014 Working Notes Proceedings of the MediaEval 2014 Workshop, Barcelona, Catalunya, Spain, October 16-17, 2014.  inproceedings URL 
Abstract: This work describes the approach proposed by the RECOD team for the Placing Task of MediaEval 2014. This task requires the definition of automatic schemes to assign ge- ographical locations to images and videos. Our approach is based on the use of as much evidences as possible (textual, visual, and/or audio descriptors) to geocode a given image/video. We estimate the location of test items by clustering the geographic coordinates of top-ranked items in one or more ranked lists defined in terms of different criteria.

BibTeX:
@inproceedings{Li2014MediaEval,
  author = {Lin Tzy Li and Otávio Augusto Bizetto Penatti and Jurandy Almeida and Giovani Chiachia and Rodrigo Tripodi Calumby and Pedro Ribeiro Mendes-Junior and Daniel Carlos Guimarães Pedronette and Ricardo da Silva Torres},
  title = {Multimedia Geocoding: The RECOD 2014 Approach},
  booktitle = {Working Notes Proceedings of the MediaEval 2014 Workshop, Barcelona, Catalunya, Spain, October 16-17, 2014.},
  year = {2014},
  url = {http://ceur-ws.org/Vol-1263/mediaeval2014_submission_81.pdf}
}
      


Li, L.T., Penatti, O.A., Fox, E.A. and Torres, R.d.S. Domain-specific Image Geocoding: A Case Study on Virginia Tech Building Photos 2013 Proceedings of the 13th ACM/IEEE-CS Joint Conference on Digital Libraries, pp. 363-366  inproceedings DOI URL 
Abstract: The use of map-based browser services is of great relevance in numerous digital libraries. The implementation of such services, however, demands the use of geocoded data collections. This paper investigates the use of image content local representations in geocoding tasks. Performed experiments demonstrate that some of the evaluated descriptors yield effective results in the task of geocoding VT building photos. This study is the first step to geocode multimedia material related to the VT April 16, 2007 school shooting tragedy.

BibTeX:
@inproceedings{Li2013JCDL,
  author = {Li, Lin Tzy and Penatti, Otávio A.B. and Fox, Edward A. and Torres, Ricardo da S.},
  title = {Domain-specific Image Geocoding: A Case Study on Virginia Tech Building Photos},
  booktitle = {Proceedings of the 13th ACM/IEEE-CS Joint Conference on Digital Libraries},
  publisher = {ACM},
  year = {2013},
  pages = {363--366},
  url = {http://doi.acm.org/10.1145/2467696.2467727},
  doi = {http://doi.org/10.1145/2467696.2467727}
}
      


Li, L.T., Almeida, J., Penatti, O.A.B., Calumby, R.T., Pedronette, D.C.G., calves , M.A.G. and da , R. Multimodal Image Geocoding: The 2013 RECOD's Approach 2013
Vol. 1043Multimedia Benchmark Workshop (Mediaeval) 
inproceedings URL 
Abstract: This work describes the approach used by the RECOD team in the MediaEval Placing Task of 2013, in which we were required to develop an automatic scheme to assign geographical locations to images. Our approach is multimodal, considering textual and visual descriptors, which are combined by a rank aggregation strategy. We estimate the location of test images based on the coordinates of top-ranked images in the list of combined results.

BibTeX:
@inproceedings{Li2013MediaEval,
  author = {Lin Tzy Li and Jurandy Almeida and Otávio A. B. Penatti and Rodrigo T. Calumby and Daniel C. G. Pedronette and Marcos A. Gon¸ calves and Ricardo da},
  title = {Multimodal Image Geocoding: The 2013 RECOD's Approach},
  booktitle = {Multimedia Benchmark Workshop (Mediaeval)},
  year = {2013},
  volume = {1043},
  url = {http://ceur-ws.org/Vol-1043/mediaeval2013submission65.pdf}
}
      


Pedronette, D.C.G. and da S. Torres, R. Exploiting Contextual Information for Image Re-Ranking and Rank Aggregation in Image Retrieval Tasks 2013 XXVI Concurso de Teses e Dissertações, XXXIII Congresso da Sociedade Brasileira de Computação, pp. -  inproceedings  
Abstract: In Content-based Image Retrieval (CBIR) systems, accurately ranking collection images is of great relevance, since users are interested in the returned images placed at the first positions. However, CBIR systems often con- sider only pairwise image analysis, i.e., they compute the similarity measures by taking into account only pairs of images, ignoring the rich information encoded in the relations among several images. On the other hand, the user perception usually considers the query specification and responses in a given context. This PhD work proposed five novel re-ranking and rank aggregation algorithms aiming at exploiting contextual information for improving the effectiveness of CBIR systems. We also proposed approaches for combining re-ranking and rank aggregation methods and for executing them efficiently on GPUs.

BibTeX:
@inproceedings{Pedronette2013CTD-SBC,
  author = {Daniel C. G. Pedronette and Ricardo da S. Torres},
  title = {Exploiting Contextual Information for Image Re-Ranking and Rank Aggregation in Image Retrieval Tasks},
  booktitle = {XXVI Concurso de Teses e Dissertações, XXXIII Congresso da Sociedade Brasileira de Computação},
  year = {2013},
  pages = {--}
}
      


Penatti, O.A.B., da S. Torres, R. and Valle, E. Image and Video Representations based on Visual Dictionaries 2013 Workshop of Thesis and Dissertations, 26th SIBGRAPI, The Conference on Graphics, Patterns, and Images, pp. -  inproceedings  
Abstract: The thesis explores three research topics involving the popular approach used for representing visual content: the visual dictionaries. The first topic concerns the generality of visual dictionaries: does a dictionary based on one dataset generalize to another dataset? Our findings create the opportunity to greatly alleviate the burden in generating dictionaries. The second topic is related to the importance of the spatial information of visual words in the image space for distinguishing types of scenes and objects. We propose an efficient and effective spatial pooling approach which presents promising results for image retrieval. And the third topic refers to the semantic information in the visual dictionary model. We claim that a bag-of-prototypes model, where the prototypes are visual words carrying semantics, is promising for improving image and video representations. Employing this model, we propose a semantically enriched dictionary based on scenes, which was effectively used for video geocoding. Defended in November 29th, 2012, the thesis has already generated 6 publications, including a best paper award. One of the proposed approaches has also obtained one of the best results in the Placing Task of MediaEval challenge in the last two years.

BibTeX:
@inproceedings{Penatti2013WTD-SIBGRAPI,
  author = {Otávio A. B. Penatti and Ricardo da S. Torres and Eduardo Valle},
  title = {Image and Video Representations based on Visual Dictionaries},
  booktitle = {Workshop of Thesis and Dissertations, 26th SIBGRAPI, The Conference on Graphics, Patterns, and Images},
  year = {2013},
  pages = {--}
}
      


Pisani, F., Pedronette, D.C.G., da Silva Torres, R. and Borin, E. Contextual Spaces Algorithm Acceleration on APUs 2013 IV Escola Regional de Alto Desempenho de São Paulo (ERAD)  inproceedings  
Abstract: Content-based image retrieval (CBIR) is a technology that helps to search images by means of their visual content. Recently, re-ranking algorithms have been proposed to exploit contextual information aiming at improving the effectiveness of CBIR systems. The image re-ranking algorithms typically demand a huge amount of computational power, which may render it prohibitive on very large data sets. In order to mitigate this problem, we propose using the computational power of Graphics Processing Units (GPU) to speedup the computation of image re-ranking algorithms. Preliminary experimental results demonstrate that relevant performance gains (up to 3.5x) can be obtained.

BibTeX:
@inproceedings{Pisani2013ERAD,
  author = {Flávia Pisani and Daniel Carlos Guimarães Pedronette and Ricardo da Silva Torres and Edson Borin},
  title = {Contextual Spaces Algorithm Acceleration on APUs},
  booktitle = {IV Escola Regional de Alto Desempenho de São Paulo (ERAD)},
  year = {2013}
}
      


Santos, J.A., Falcão, A.X., da S. Torres, R., Gosselin, P.-H. and Philipp-Foliguet, S. Semi-automatic Classification of Remote Sensing Images 2013 Workshop of Thesis and Dissertations, 26th SIBGRAPI, The Conference on Graphics, Patterns, and Images, pp. -  inproceedings  
Abstract: With the increasing accessibility to new technologies, the main problems in region recognition of remote sensing images are: (1) classification methods are dependent on the segmentation quality; and (2) the selection of representative samples for training. The major challenge is that the samples indicated by the user are not always enough to define the best segmentation scale. Furthermore, the indication of samples can be costly, since it often requires to visit studied places in loco. This PhD thesis1 addressed the segmentation-dependence problem by introducing two new approaches that rely on multiple scales instead of using only one segmentation result. The selection of representative samples, on the other hand, was supported in this work by the development of a new interactive classification approach based on active learning. Significant contributions were also obtained concerning the description of regions in remote sensing images by means of: an evaluation study of 19 descriptors; and two new strategies for speeding up feature extraction from a hierarchy of segmented regions.

BibTeX:
@inproceedings{Santos2013WTD-SIBGRAPI,
  author = {Jefersson A. Santos and Alexandre X. Falcão and Ricardo da S. Torres and Gosselin, P.-H. and Sylvie Philipp-Foliguet},
  title = {Semi-automatic Classification of Remote Sensing Images},
  booktitle = {Workshop of Thesis and Dissertations, 26th SIBGRAPI, The Conference on Graphics, Patterns, and Images},
  year = {2013},
  pages = {--}
}
      


Alberton, B., Almeida, J., Hanneken, R., da S. Torres, R., Menzel, A. and Morellato, L.P. Remote Phenology: Applying digital images to monitor leaf phenology in a Brazilian Cerrado Savanna 2012 49th Annual Meeting of the Association for Tropical Biology and Conservation  inproceedings  
Abstract: Plant phenology has gained importance in the context of global change research, stimulating the development of new technologies for phenological observation. Digital cameras have been successfully used as multi-channel imaging sensors, providing measures of leaf color change information (RGB channels), or leafing phenological changes in plants. We monitored leaf-changing patterns of a cerrado-savanna vegetation by taken daily digital images. We extract RGB channels from digital images and correlated with phenological changes. Our first goals were: i) to test if there is any statistical difference between time series from different hours of the day, ii) to test if time series from different regions of interest (ROI) in the image differ from a combination of ROIs or the total image. A digital hemispherical lens camera was setup at the top of a 18 m high tower to automatically take a daily sequence of five JPEG images per hour, from 6:00 to 18:00h. The present study was based on the analysis of over 2,700 images, recorded between August 28th and October 3rd, 2011. The image analysis was conducted by defining ten ROIs in the original digital image, including total or partial images, and six plant species. We calculated the relative values of green (% of Green) of each image and tested for correlations using ZNCC metric. Our preliminary results indicate that: (i) midday hours are the best for cerrado-savanna remote phenology, (ii) plant species ROIs are more sensitive to leaf changes than the combination of several ROIs or community images, and (iii) differences among species' ROIs may be related to its particular leaf change strategy.

BibTeX:
@inproceedings{Alberton2012Poster-ATBC,
  author = {Bruna Alberton and Jurandy Almeida and Raimund Hanneken and Ricardo da S. Torres and Annette Menzel and Leonor Patrícia Morellato},
  title = {Remote Phenology: Applying digital images to monitor leaf phenology in a Brazilian Cerrado Savanna},
  booktitle = {49th Annual Meeting of the Association for Tropical Biology and Conservation},
  year = {2012}
}
      


Alberton, B., Almeida, J., Hanneken, R., da S. Torres, R., Menzel, A. and Morellato, P. Remote phenology: applying digital images to monitor leaf phenology in a Brazilian cerrado savanna 2012 Phenology 2012 Conference  inproceedings  
Abstract: Plant phenology has gained importance as the simplest and most reliable indicator of responses in the context of global change research, stimulating the development of new technologies for phenological observation. Digital cameras have been successfully used as multi-channel imaging sensors, providing measures of leaf color change information in plants, using red green and blue (RGB) channels of digital images. We monitored leaf-changing patterns of Brazilian cerrado-savanna vegetation taking daily digital images. We extracted RGB channels from digital images and correlated them with phenological changes. We ask: i) whether time series significantly differ between different hours of the day, and ii) whether time series from different regions of interest (ROI) in the image (e.g the canopy of one tree) differ from a combination of ROIs or the total image...

BibTeX:
@inproceedings{Alberton2012Poster-Phenology,
  author = {Bruna Alberton and Jurandy Almeida and Raimund Hanneken and Ricardo da S. Torres and Annette Menzel and Patrícia Morellato},
  title = {Remote phenology: applying digital images to monitor leaf phenology in a Brazilian cerrado savanna},
  booktitle = {Phenology 2012 Conference},
  year = {2012}
}
      


Almeida, J., Leite, N.J. and da S. Torres, R. Recuperação de Vídeos Comprimidos por Conteúdo 2012 XXV Concurso de Teses e Dissertações, XXXII Congresso da Sociedade Brasileira de Computação, pp. -  inproceedings  
Abstract: Most of existing systems for video retrieval rely on algorithms and methods which are computationally expensive, limiting their scope of application. Contrary to this trend, the market has shown a growing demand for mobile and embedded devices. In this context, this thesis introduces five novel approa- ches for the analysis, indexing, and retrieval of digital videos in limited capacity devices. All these contributions are combined to create a computationally fast system, which is able to achieve a quality level superior to current solutions.

BibTeX:
@inproceedings{Almeida2012CTD-SBC,
  author = {Jurandy Almeida and Neucimar J. Leite and Ricardo da S. Torres},
  title = {Recuperação de Vídeos Comprimidos por Conteúdo},
  booktitle = {XXV Concurso de Teses e Dissertações, XXXII Congresso da Sociedade Brasileira de Computação},
  year = {2012},
  pages = {--}
}
      


Almeida, J., Salles, T., Martins, E.F., Penatti, O.A.B., da S. Torres, R., Gonçalves, M.A. and Almeida, J.M. UNICAMP-UFMG at MediaEval 2012: Genre Tagging Task 2012 MediaEval  inproceedings DOI  
Abstract: Developed in the context of Genre Tagging Task at Media- Eval 2012, this work consists in automatically assigning genre tags to a set of Internet videos. We approach this task from the classification point of view and focus on different learn- ing strategies: video similarity for processing visual content and an ensemble of classifiers for text-processing. In this paper, we describe the proposed framework and report the results obtained on the official submission runs.

BibTeX:
@inproceedings{Almeida2012MediaEval,
  author = {Jurandy Almeida and Thiago Salles and Eder Ferreira Martins and Otávio Augusto Bizetto Penatti and Ricardo da S. Torres and Marcos André Gonçalves and Jussara M. Almeida},
  title = {UNICAMP-UFMG at MediaEval 2012: Genre Tagging Task},
  booktitle = {MediaEval},
  year = {2012},
  doi = {http://ceur-ws.org/Vol-927/mediaeval2012\_submission\_13.pdf}
}
      


Almeida, J., Leite, N.J. and da S. Torres, R. Content-Based Retrieval of Compressed Videos 2012 Workshop of Theses and Dissertations, Conference on Graphics, Patterns and Images (25th SIBGRAPI), pp. -  inproceedings  
Abstract: The fast evolution of technology has led to a growing demand for video data, increasing the amount of research into efficient systems to manage video material. Most of existing systems for video retrieval rely on algorithms and methods which are computationally expensive, in terms of both time and space, limiting their application to the academic world or big companies. Contrary to this trend, the market has shown a growing demand for mobile and embedded devices. In this scenario, it is imperative the development of effective and efficient techniques in order to make such technologies available to a larger public. In this context, this PhD thesis introduces five novel approaches for the analysis, indexing, and retrieval of digital videos. All these contributions are combined to create a computationally fast system for content-based video retrieval, which is able to achieve a quality level superior to current solutions.

BibTeX:
@inproceedings{Almeida2012WTD-SIBGRAPI,
  author = {Jurandy Almeida and Neucimar J. Leite and Ricardo da S. Torres},
  title = {Content-Based Retrieval of Compressed Videos},
  booktitle = {Workshop of Theses and Dissertations, Conference on Graphics, Patterns and Images (25th SIBGRAPI)},
  year = {2012},
  pages = {--}
}
      


Li, L.T., Almeida, J., Pedronette, D.C.G., Penatti, O.A.B. and da S. Torres, R. A Multimodal Approach for Video Geocoding 2012 MediaEval  inproceedings DOI  
Abstract: Developed in the context of placing task at MediaEval 2012, this work addresses the problem of automatically assigning geographical coordinates to videos. This year our group extended the implementation of our framework for multimodal geocoding for combining textual and visual descriptors. In this paper, we describe our approach and report the results for 2012 datasets.

BibTeX:
@inproceedings{Li2012MediaEval,
  author = {Lin Tzy Li and Jurandy Almeida and Daniel Carlos Guimarães Pedronette and Otávio Augusto Bizetto Penatti and Ricardo da S. Torres},
  title = {A Multimodal Approach for Video Geocoding},
  booktitle = {MediaEval},
  year = {2012},
  doi = {http://ceur-ws.org/Vol-927/mediaeval2012\_submission\_19.pdf}
}
      


Li, L.T., Pedronette, D.C.G., Almeida, J., Penatti, O.A.B., Calumby, R.T. and da S. Torres, R. Multimedia Multimodal Geocoding 2012 20th ACM SIGSPATIAL International Conference on Advances in Geographic Information Systems (ACM SIGSPATIAL GIS)  inproceedings DOI  
Abstract: This work is developed in the context of the placing task of the MediaEval 2011 initiative. The objective is to geocode (or geotag) a set of videos, i.e., automatically assign geographical coordinates to them. This paper presents an architecture for multimodal geocoding that exploits both visual and textual descriptions associated with videos. This work also describes our efforts regarding the implementation of this architecture to demonstrate its applicability. Conducted experiments show how our multimodal approach enhances the results compared to relying on a single modality.

BibTeX:
@inproceedings{Lin2012ACMGIS,
  author = {Lin Tzy Li and Daniel Carlos Guimarães Pedronette and Jurandy Almeida and Otávio Augusto Bizetto Penatti and Rodrigo Tripodi Calumby and Ricardo da S. Torres},
  title = {Multimedia Multimodal Geocoding},
  booktitle = {20th ACM SIGSPATIAL International Conference on Advances in Geographic Information Systems (ACM SIGSPATIAL GIS)},
  year = {2012},
  doi = {http://doi.org/10.1145/2424321.2424393}
}
      


Pedronette, D.C.G., Borin, E., da Silva Torres, R. and Jr., M.B. Efficient Computation of Image Re-Ranking using GPUs 2012 III Escola Regional de Alto Desempenho de São Paulo (ERAD)  inproceedings  
Abstract: Content-based image retrieval (CBIR) is a technology that helps to search images by means of their visual content. Recently, re-ranking algorithms have been proposed to exploit contextual information aiming at improving the effectiveness of CBIR systems. The image re-ranking algorithms typically demand a huge amount of computational power, which may render it prohibitive on very large data sets. In order to mitigate this problem, we propose using the computational power of Graphics Processing Units (GPU) to speedup the computation of image re-ranking algorithms. Preliminary experimental results demonstrate that relevant performance gains (up to 3.5x) can be obtained.

BibTeX:
@inproceedings{Pedronette2012ERAD,
  author = {Daniel Carlos Guimarães Pedronette and Edson Borin and Ricardo da Silva Torres and Mauricio Breternitz Jr.},
  title = {Efficient Computation of Image Re-Ranking using GPUs},
  booktitle = {III Escola Regional de Alto Desempenho de São Paulo (ERAD)},
  year = {2012}
}
      


Pedronette, D.C.G. and da S. Torres, R. Exploiting Contextual Information in Image Retrieval Tasks 2012 Workshop of Theses and Dissertations, Conference on Graphics, Patterns and Images (25th SIBGRAPI), pp. -  inproceedings  
Abstract: In Content-based Image Retrieval (CBIR) systems, accurately ranking images is of great relevance, since users are interested in the returned images placed at the first positions, which usually are the most relevant ones. In general, CBIR systems consider only pairwise image analysis, that is, compute similarity measures considering only pairs of images, ignoring the rich information encoded in the relations among several images. On the other hand, the user perception usually considers the query specification and responses in a given context. We propose five re-ranking and rank aggregation algorithms aiming at exploit contextual information for improving the effectiveness of CBIR systems. Re-ranking algorithms exploit contextual in- formation, encoded in the relationships among collection images, while rank aggregation approaches have been used to combine results produced by different image descriptors. We also propose approaches for combining the re-ranking and rank aggregation methods and for efficient re-ranking computation on GPUs.

BibTeX:
@inproceedings{Pedronette2012WTD-SIBGRAPI,
  author = {Daniel C. G. Pedronette and Ricardo da S. Torres},
  title = {Exploiting Contextual Information in Image Retrieval Tasks},
  booktitle = {Workshop of Theses and Dissertations, Conference on Graphics, Patterns and Images (25th SIBGRAPI)},
  year = {2012},
  pages = {--}
}
      


Almeida, J., Pinto-Cáceres, S., da S. Torres, R. and Leite, N.J. Intuitive Video Browsing Along Hierarchical Trees 2011 II Simpósio de Processamento de Sinais da UNICAMP (SPS)  inproceedings  
Abstract: Recent advances in technology have increased the availability of video data, creating a strong requirement for efficient systems to manage those materials. Making efficient use of video information requires that data be accessed in a user-friendly way. Ideally, one would like to perform video search using an intuitive tool. Most of existing browsers for the interactive search of video sequences, however, have employed a too rigid layout to arrange the results, restricting users to explore the results using list- or grid-based layouts. In this paper, we present a novel approach for the interactive search that displays the result set in a flexible manner. The proposed method is based on a hierarchical structure called Divisive-Agglomerative Hierarchical Clustering (DAHC). It is able to group together videos with similar content and to organize the result set over a well-defined visual structure called Clustering Set. This strategy makes the navigation more coherent and engaging to users.

BibTeX:
@inproceedings{Almeida2011SPS,
  author = {Jurandy Almeida and Sheila Pinto-Cáceres and Ricardo da S. Torres and Neucimar J. Leite},
  title = {Intuitive Video Browsing Along Hierarchical Trees},
  booktitle = {II Simpósio de Processamento de Sinais da UNICAMP (SPS)},
  year = {2011}
}
      


Faria, F.A., Calumby, R.T. and da S. Torres, R. RECOD at ImageCLEF 2011: Medical Modality Classification using Genetic Programming 2011 CLEF (Notebook Papers/Labs/Workshop)  inproceedings  
Abstract: This paper describes the participation of the RECOD group on the ImageCLEF 2011 Medical Modality Classification sub-task. We present an approach based on genetic programming and kNN for image classification. In our approach the genetic programming is used for the learning of good functions for the combination of similarities obtained from a set of global descriptors for different visual evidences such as color, texture, and shape. For each class of the dataset a combination function was learned and used as a kNN classifier. Final classification results were generated by a majority voting scheme with the voting functions from each class. Preliminary experiments have shown a good effectiveness of the approach and its potential for improvements.

BibTeX:
@inproceedings{Faria2011CLEF,
  author = {Fabio Augusto Faria and Rodrigo Tripodi Calumby and Ricardo da S. Torres},
  title = {RECOD at ImageCLEF 2011: Medical Modality Classification using Genetic Programming},
  booktitle = {CLEF (Notebook Papers/Labs/Workshop)},
  year = {2011}
}
      


Faria, F.A., Calumby, R.T., Veloso, A., Rocha, A. and da S. Torres, R. Uso de Técnicas de Aprendizagem para Classificação e Recuperação de Imagens 2011 Workshop of Theses and Dissertations, 24th Conference on Graphics, Patterns and Images (SIBGRAPI), pp. 174-179  inproceedings  
Abstract: This work sought to use learning techniques to combine visual evidences (e.g., color, texture, and shape) to obtain effectiveness improvements for image classification and retrieval tasks. The main contributions of this work are: design and development of a classifier using Genetic Programming techniques to combine visual evidences to be used in image classification tasks as well as an approach for content-based image retrieval based on association rules.

BibTeX:
@inproceedings{Faria2011WTD-SIBGRAPI,
  author = {Fabio Augusto Faria and Rodrigo T. Calumby and Adriano Veloso and Anderson Rocha and Ricardo da S. Torres},
  title = {Uso de Técnicas de Aprendizagem para Classificação e Recuperação de Imagens},
  booktitle = {Workshop of Theses and Dissertations, 24th Conference on Graphics, Patterns and Images (SIBGRAPI)},
  year = {2011},
  pages = {174--179}
}
      


Li, L.T., Almeida, J. and da S. Torres, R. RECOD Working Notes for Placing Task MediaEval 2011 2011 MediaEval  inproceedings URL 
Abstract: This work is developed in the context of placing task at MediaEval 2011. It consists in automatically assigning geo- graphical coordinates to a set of videos. Our group proposed an architecture design for the multimodal geocoding. In this paper, we focused on implementing a simple content-based approach, which is part of the proposed framework. The reported results show our strategy compared to those from previous year participant using only visual content to accomplish this task.

BibTeX:
@inproceedings{Li2011MediaEval,
  author = {Lin Tzy Li and Jurandy Almeida and Ricardo da S. Torres},
  title = {RECOD Working Notes for Placing Task MediaEval 2011},
  booktitle = {MediaEval},
  year = {2011},
  url = {http://ceur-ws.org/Vol-807/Li_UNICAMP_Placing_me11wn.pdf}
}
      


Yang, S., Kavanaugh, A.L., Kozievitch, N.P., Li, L.T., Srinivasan, V., Sheetz, S.D., Whalen, T., Shoemaker, D., da S. Torres, R. and Fox, E.A. CTRnet DL for disaster information services 2011 Joint International Conference on Digital Libraries (JCDL), pp. 437-438  inproceedings DOI  
Abstract: We describe our work in collecting, analyzing and visualizing online information (e.g., Web documents, images, tweets), which are to be maintained by the Crisis, Tragedy and Recovery Network (CTRnet) digital library. We have been collecting resources about disaster events, as well as campus and other major shooting events, in collaboration with the Internet Archive (IA). Social media data (e.g., tweets, Facebook data) also have been collected and analyzed. Analyzed results are visualized using graphs and tag clouds. Exploratory content-based image retrieval has been applied in one of our image collections. We explain our CTR ontology development methodology and collaboration with Arlington County, VA and IBM, in a Center for Community Security and Resilience funded project.

BibTeX:
@inproceedings{Yang2011Poster-JCDL,
  author = {Seungwon Yang and Andrea L. Kavanaugh and Nádia P. Kozievitch and Lin Tzy Li and Venkat Srinivasan and Steven D. Sheetz and Travis Whalen and Donald Shoemaker and Ricardo da S. Torres and Edward A. Fox},
  title = {CTRnet DL for disaster information services},
  booktitle = {Joint International Conference on Digital Libraries (JCDL)},
  year = {2011},
  pages = {437--438},
  doi = {http://doi.org/10.1145/1998076.1998173}
}
      


Almeida, J., da S. Torres, R. and Leite, N.J. BP-tree: an efficient index for similarity search in high-dimensional metric spaces 2010 19th ACM Conference on Information and Knowledge Management (CIKM), pp. 1365-1368  inproceedings DOI  
Abstract: Similarity search in high-dimensional metric spaces is a key operation in many applications, such as multimedia databases, image retrieval, object recognition, and others. The high dimensionality of the data requires special index structures to facilitate the search. Most of existing indexes are constructed by partitioning the data set using distance-based criteria. However, those methods either produce disjoint partitions, but ignore the distribution properties of the data; or produce non-disjoint groups, which greatly affect the search performance. In this paper, we study the performance of a new index structure, called Ball-and-Plane tree (BP-tree), which overcomes the above disadvantages. BP-tree is constructed by recursively dividing the data set into compact clusters. Distinctive from other techniques, it integrates the advantages of both disjoint and non-disjoint paradigms in order to achieve a structure of tight and low overlapping clusters, yielding significantly improved performance. Results obtained from an extensive experimental evaluation with real-world data sets show that BP-tree consistently outperforms state-of-the-art solutions.

BibTeX:
@inproceedings{Almeida2010CIKM,
  author = {Jurandy Almeida and Ricardo da S. Torres and Neucimar Jerônimo Leite},
  title = {BP-tree: an efficient index for similarity search in high-dimensional metric spaces},
  booktitle = {19th ACM Conference on Information and Knowledge Management (CIKM)},
  year = {2010},
  pages = {1365--1368},
  doi = {http://doi.org/10.1145/1871437.1871622}
}
      


Almeida, J., Leite, N.J. and da S. Torres, R. Recupera¸ cão de Vídeos Digitais: Métodos, Propostas e Desafios 2010 I Simpósio de Processamento de Sinais da UNICAMP (SPS), pp. 75-77  inproceedings  
Abstract: Avan¸ cos em técnicas de aquisi¸ cão, compressão e transmissão de dados têm permitido a cria¸ cão de bibliotecas de vídeo digital cada vez maiores. Na medida em que cresce o volume de vídeos digitais nessas cole¸ ces, cresce também o interesse em sistemas capazes de gerenciá-los de maneira eficiente. Essa dificuldade vem sendo resolvida por meio de sistemas capazes de abstrair representa¸ cões semânticas de alto nível a partir das informa¸ cões de baixo nível dos vídeos. Devido à complexidade desse material, existem quatro desafios principais na constru¸ cão desses sistemas: (1) implementar um algoritmo para codificar as propriedades de baixo nível dos dados de vídeo em vetores de características; (2) desenvolver uma medida de similaridade para comparar vídeos a partir de seus vetores; (3) responder rapidamente a consultas por similaridade sobre, potencialmente, milhões de sequências de vídeo; e (4) apresentar os resultados de forma amigável ao usuário. Inúmeras técnicas têm sido propostas para atender a esses requisitos. Todavia, apesar de todos os esfor¸ cos o gerenciamento de vídeos digitais ainda está em seu estado preliminar. Este artigo apresenta uma revisão de diferentes métodos propostos para atender aos quesitos supracitados, novas solu¸ cões propostas para alguns problemas abordados e questões que se encontram em aberto.

BibTeX:
@inproceedings{Almeida2010SPS,
  author = {Jurandy Almeida and Neucimar J. Leite and Ricardo da S. Torres},
  title = {Recupera¸ cão de Vídeos Digitais: Métodos, Propostas e Desafios},
  booktitle = {I Simpósio de Processamento de Sinais da UNICAMP (SPS)},
  year = {2010},
  pages = {75--77}
}
      


Andrade, F., da S. Torres, R. and Pedrini, H. Combina¸ cão de Descritores Locais e Globais para Reconhecimento de Objetos em Vídeo 2010 I Simpósio de Processamento de Sinais da UNICAMP (SPS), pp. 84-86  inproceedings  
Abstract: O processo de reconhecimento de objetos tem despertado bastante interesse de pesquisadores das áreas de visão computacional e processamento de imagens/vídeos, cujo objetivo principal é a determina¸ cãoo da identidade ou classe a qual um objeto pertence, bem como sua localiza¸ cão. Neste trabalho apresentamos uma abordagem que visa investigar diversos descritores locais e globais de imagens com aplica¸ cão em reconhecimento de objetos em sequências de vídeos, bem como a implementa¸ cão de mecanismos para realizar a combina¸ cão de descritores no mesmo cenário.

BibTeX:
@inproceedings{Andrade2010SPS,
  author = {Felipe Andrade and Ricardo da S. Torres and Hélio Pedrini},
  title = {Combina¸ cão de Descritores Locais e Globais para Reconhecimento de Objetos em Vídeo},
  booktitle = {I Simpósio de Processamento de Sinais da UNICAMP (SPS)},
  year = {2010},
  pages = {84--86}
}
      


Armigliatto, G., Valle, E. and da S. Torres, R. Anota¸ cão Automática de Imagens Usando Dicionários Visuais e Regras de Associa¸ cão 2010 I Simpósio de Processamento de Sinais da UNICAMP (SPS), pp. 29-31  inproceedings  
Abstract: Com os avan¸ cos tecnológicos, grandes cole¸ cões de imagens são geradas, manipuladas e armazenadas em bancos de imagens. Dado o grande tamanho desses bancos, verifica-se a necessidade de criar ferramentas que gerenciem de forma eficaz as imagens neles contidas. Um dos problemas desse gerenciamento é a recupera¸ cão das imagens no banco, e uma forma de fazê-lo é baseada no uso de anota¸ cões textuais associadas a imagem. Entretanto, a anota¸ cão manual das bases implica grandes custos, além de ser demorada. Este trabalho tem por objetivo explorar técnicas para anota¸ cão automática de imagens usando dicionários visuais e regras de associa¸ cão.

BibTeX:
@inproceedings{Armigliatto2010SPS,
  author = {Guilherme Armigliatto and Eduardo Valle and Ricardo da S. Torres},
  title = {Anota¸ cão Automática de Imagens Usando Dicionários Visuais e Regras de Associa¸ cão},
  booktitle = {I Simpósio de Processamento de Sinais da UNICAMP (SPS)},
  year = {2010},
  pages = {29--31}
}
      


Bueno, L.M., Valle, E. and da S. Torres, R. Estudo Comparativo de Descritores Locais de Imagens Aplicado a Grandes Bases de Dados 2010 I Simpósio de Processamento de Sinais da UNICAMP (SPS), pp. 47-48  inproceedings  
Abstract: Descritores locais de imagens são amplamente utilizados em diversas aplica¸ cões de reconhecimento de objetos ou cenas, bem como em recupera¸ cão de semi-réplicas. Muitos descritores locais foram propostos na literatura para caracterizar regiões em torno de pontos de interesse em imagens. Pontos de interesse em imagens são determinados por detectores. Apesar das pesquisas em compara¸ cão entre descritores locais estarem avan¸ cadas, pouco foi feito quando à avalia¸ cão deles para grandes bases de imagens. O objetivo desse trabalho é fazer um estudo comparativo entre descritores locais com enfoque na escalabilidade deles em rela¸ cão ao tamanho da base.

BibTeX:
@inproceedings{Bueno2010SPS,
  author = {Lucas M. Bueno and Eduardo Valle and Ricardo da S. Torres},
  title = {Estudo Comparativo de Descritores Locais de Imagens Aplicado a Grandes Bases de Dados},
  booktitle = {I Simpósio de Processamento de Sinais da UNICAMP (SPS)},
  year = {2010},
  pages = {47--48}
}
      


Calumby, R.T. and da S. Torres, R. Realimenta¸ cão de Relevância para Recupera¸ cão Visual e Textual de Imagens 2010 I Simpósio de Processamento de Sinais da UNICAMP (SPS), pp. 32-34  inproceedings  
Abstract: Este trabalho apresenta uma abordagem para recupera¸ cão multimodal de imagens com realimenta¸ cão de relevância baseada em programa¸ cão genética. Supõe-se que cada imagem da cole¸ cão possui informa¸ cão textual associada, além de ter suas propriedades visuais codificadas em vetores de características. A partir da informa¸ cão obtida ao longo das itera¸ cões de realimenta¸ cão de relevância, programa¸ cão genética é utilizada para a cria¸ cão de fun¸ cões de combina¸ cão de medidas de similaridades eficazes. Com essas novas fun¸ cões, valores de similaridades diversos são combinados em uma única medida, que mais adequadamente reflete as necessidades do usuário. A abordagem proposta mostrou melhores resultados para recupera¸ cão multimodal frente a utiliza¸ c ao das modalidades isoladas. Além disso, foram obtidos resultados para recupera¸ cão multimodal superiores as melhores submissões para a ImageCLEF Photographic Retrieval Task 2008.

BibTeX:
@inproceedings{Calumby2010SPS,
  author = {Rodrigo Tripodi Calumby and Ricardo da S. Torres},
  title = {Realimenta¸ cão de Relevância para Recupera¸ cão Visual e Textual de Imagens},
  booktitle = {I Simpósio de Processamento de Sinais da UNICAMP (SPS)},
  year = {2010},
  pages = {32--34}
}
      


Faria, F., Rocha, A. and da S. Torres, R. Classificação de Imagens usando Programação Genética 2010 I Simpósio de Processamento de Sinais da UNICAMP (SPS), pp. 26-28  inproceedings  
Abstract: Técnicas de aprendizagem de máquina vêm sendo empregadas em diversas áreas do conhecimento (medicina, biologia, seguran¸ ca, entre outras). Este trabalho, busca mostrar o uso da técnica de Programa¸ cão Genética (PG) em tarefas de classifica¸ cão de imagens. PG é uma técnica da inteligência artificial que busca solu¸ cões ótimas inspirada pela teoria de sele¸ cão natural das espécies. Indivíduos mais aptos (melhores solu¸ cões) tendem a evoluir e se reproduzir nas gera¸ cões futuras. Este artigo discute como a PG pode ser utilizada para combinar diferentes evidências visuais psara melhorar o desempenho em tarefas de classifica¸ cão de imagens.

BibTeX:
@inproceedings{Faria2010SPS,
  author = {Fabio Faria and Anderson Rocha and Ricardo da S. Torres},
  title = {Classificação de Imagens usando Programação Genética},
  booktitle = {I Simpósio de Processamento de Sinais da UNICAMP (SPS)},
  year = {2010},
  pages = {26--28}
}
      


Kozievitch, N.P., da S. Torres, R., Andrade, F.S.P., Murthy, U., Fox, E.A. and Hallerman, E. A Teaching Tool for Parasitology: Enhancing Learning with Annotation and Image Retrieval 2010 esearch and Advanced Technology for Digital Libraries, 14th European Conference (ECDL), pp. 466-469  inproceedings DOI  
Abstract: Parasitology is a basic course in life sciences curricula, but up to now it has few computer-assisted teaching tools. We present SuperIDR, a tool which supports annotation and search (based on a textual and a visual description) in the biodiversity domain. In addition, it provides a feature to aid comparison of morphological characteristics among different species. Preliminary results with two experiments show that students found the tool to be very useful, contributing to an alternative learning approach.

BibTeX:
@inproceedings{Kozievitch2010ECDL,
  author = {Nádia P. Kozievitch and Ricardo da S. Torres and Felipe S. P. Andrade and Uma Murthy and Edward A. Fox and Eric Hallerman},
  title = {A Teaching Tool for Parasitology: Enhancing Learning with Annotation and Image Retrieval},
  booktitle = {esearch and Advanced Technology for Digital Libraries, 14th European Conference (ECDL)},
  year = {2010},
  pages = {466--469},
  doi = {http://doi.org/10.1007/978-3-642-15464-5%5C_58}
}
      


Kozievitch, N.P. and da S. Torres, R. Describing OAI-ORE from the 5S Framework Perspective 2010 The Role of Digital Libraries in a Time of Global Change, 12th International Conference on Asia-Pacific Digital Libraries (ICADL), pp. 260-261  inproceedings DOI  
Abstract: Despite the popularity of applications which manage complex digital objects, few attempts have been made to formally characterize them and their services. This poster addresses this problem starting an analysis of OAI-ORE specifications under the 5s framework perspective, verifying how they can be integrated to describe complex digital objects as resources that could be later exchanged.

BibTeX:
@inproceedings{Kozievitch2010Poster-ICADL,
  author = {Nádia P. Kozievitch and Ricardo da S. Torres},
  title = {Describing OAI-ORE from the 5S Framework Perspective},
  booktitle = {The Role of Digital Libraries in a Time of Global Change, 12th International Conference on Asia-Pacific Digital Libraries (ICADL)},
  year = {2010},
  pages = {260--261},
  doi = {http://doi.org/10.1007/978-3-642-13654-2%5C_34}
}
      


Kozievitch, N., da S. Torres, R., ao , T.R.C.F., Ramos, E.J., Andrade, F., Allegretti, S.M., Ueta, M.T., Madi, R.R., Murthy, U., Fox, E.A., Chen, Y. and Hallerm, E.M. A Geographic Annotation Service in SuperIDR 2010 GIS and Remote Sensing Research Symposium (VTGIS)  inproceedings  
BibTeX:
@inproceedings{Kozievitch2010Poster-VTGIS,
  author = {Nádia Kozievitch and Ricardo da S. Torres and Tiago R. C. Falcão and Evandro J. Ramos and Felipe Andrade and Silmara Marques Allegretti and Marlene Tiduko Ueta and Rubens Riscala Madi and Uma Murthy and Edward A. Fox and Yinlin Chen and Eric M. Hallerm},
  title = {A Geographic Annotation Service in SuperIDR},
  booktitle = {GIS and Remote Sensing Research Symposium (VTGIS)},
  year = {2010}
}
      


Muraro, E. and da S. Torres, R. Busca Baseada em Ontologias de Objetos Digitais Armazenados em Repositórios na Web 2010 IX Workshop de Teses e Dissertações em Bancos de Dados, XXV Brazilian Symposium on Databases  inproceedings  
Abstract: O problema-alvo desta pesquisa de mestrado consiste na especifica¸ cão e desenvolvimento de um mecanismo de busca baseado em ontologias, visando à recupera¸ cão de objetos digitais armazenados em repositórios na Web. Esses objetos digitais podem ser tanto dados (textos, vídeos e imagens), quanto software (aplicativos e servi¸ cos). Estes objetos podem ser representados utilizando-se componentes de conteúdos de dados. Os componentes utilizam recursos da Web Semântica, como metadados e ontologias, na sua constru¸ cão e implementa¸ cão. Nesta pesquisa, em particular, serão utilizados componentes implementados a com a estrutura Digital Content Component (DCC). O uso de DCC para encapsulamento de dados e software foi proposto recentemente no Instituto de Computa¸ cão da Unicamp. Basicamente, um DCC é composto por: conteúdo (dado ou software), um XML com a especifica¸ cão da estrutura interna do componente, uma representa¸ cão semântica da interface do componente e metadados para representa¸ cão do componente. O servi¸ co de busca de DCC será validado no cenário de biblioteca digital de artigos científicos. Nesta biblioteca os artigos e seus componentes (figuras, tabelas, etc) serão encapsulados em DCCs.

BibTeX:
@inproceedings{Muraro2010WTDBD-SBBD,
  author = {Emerson Muraro and Ricardo da S. Torres},
  title = {Busca Baseada em Ontologias de Objetos Digitais Armazenados em Repositórios na Web},
  booktitle = {IX Workshop de Teses e Dissertações em Bancos de Dados, XXV Brazilian Symposium on Databases},
  year = {2010}
}
      


Pedronette, D. and da S. Torres, R. Exploiting Shape Matching and Re-Ranking Strategies for Content-Based Image Retrieval Tasks 2010 I Simpósio de Processamento de Sinais da UNICAMP (SPS), pp. 44-46  inproceedings  
Abstract: This paper describes ongoing research concerning the proposal of new methods methods proposed aiming to improve the effectiveness of Content-Based Image Retrieval (CBIR) systems. Conducted experiments on MPEG-7 database demonstrates the effectiveness of our methods.

BibTeX:
@inproceedings{Pedronette2010SPS,
  author = {Daniel Pedronette and Ricardo da S. Torres},
  title = {Exploiting Shape Matching and Re-Ranking Strategies for Content-Based Image Retrieval Tasks},
  booktitle = {I Simpósio de Processamento de Sinais da UNICAMP (SPS)},
  year = {2010},
  pages = {44--46}
}
      


Penatti, O.A.B. and da S. Torres, R. Eva: an evaluation tool for comparing descriptors in content-based image retrieval tasks 2010 11th ACM SIGMM International Conference on Multimedia Information Retrieval (MIR), pp. 413-416  inproceedings DOI  
Abstract: This paper presents Eva, a tool for evaluating image descriptors for content-based image retrieval. Eva integrates the most common stages of an image retrieval process and provides functionalities to facilitate the comparison of image descriptors in the context of content-based image retrieval. Eva supports the management of image descriptors and image collections and creates a standardized environment to run comparative experiments using them.

BibTeX:
@inproceedings{Penatti2010Demo-MIR,
  author = {Otávio A. B. Penatti and Ricardo da S. Torres},
  title = {Eva: an evaluation tool for comparing descriptors in content-based image retrieval tasks},
  booktitle = {11th ACM SIGMM International Conference on Multimedia Information Retrieval (MIR)},
  year = {2010},
  pages = {413--416},
  doi = {http://doi.org/10.1145/1743384.1743455}
}
      


Penatti, O.A.B. and da S. Torres, R. User-Oriented Evaluation of Color Descriptors for Web Image Retrieval 2010 esearch and Advanced Technology for Digital Libraries, 14th European Conference (ECDL), pp. 486-489  inproceedings DOI  
Abstract: This paper proposes a methodology for effectiveness evaluation in content-based image retrieval systems. The methodology is based on the opinion of real users. This paper also presents the results of using this methodology to evaluate color descriptors for Web image retrieval. The experiments were performed using a database containing more than 230 thousand heterogeneous images that represents the existing content on the Web.

BibTeX:
@inproceedings{Penatti2010ECDL,
  author = {Otávio A. B. Penatti and Ricardo da S. Torres},
  title = {User-Oriented Evaluation of Color Descriptors for Web Image Retrieval},
  booktitle = {esearch and Advanced Technology for Digital Libraries, 14th European Conference (ECDL)},
  year = {2010},
  pages = {486--489},
  doi = {http://doi.org/10.1007/978-3-642-15464-5%5C_63}
}
      


Penatti, O.A. and da S. Torres, R. Explorando Dicionários Visuais em Buscas de Imagens na Web 2010 I Simpósio de Processamento de Sinais da UNICAMP (SPS), pp. 35-37  inproceedings  
Abstract: O uso de dicionários visuais vem crescendo nos últimos anos e vem se mostrando atrativo para muitas aplica¸ cões. Seus principais benefícios estão relacionados à capacidade de entender o conteúdo visual de imagens e vídeos, como por exemplo, indicar e classificar os objetos existentes em uma imagem. Este entendimento ocorre por meio da associa¸ cão automática de conceitos semânticos aos padrões visuais dos dados. Este projeto visa especificar e implementar um sistema de busca de imagens na Web explorando o uso dicionários visuais. As principais contribui¸ cões deste trabalho são: estudo comparativo das abordagens atuais para a gera¸ cão de dicionários visuais; especifica¸ cão e implementa¸ cão de novas técnicas para a gera¸ cão de dicionários visuais; especifica¸ cão e implementa¸ cão parcial de sistema de busca de imagens na Web baseado em dicionários visuais.

BibTeX:
@inproceedings{Penatti2010SPS,
  author = {Otávio A. Penatti and Ricardo da S. Torres},
  title = {Explorando Dicionários Visuais em Buscas de Imagens na Web},
  booktitle = {I Simpósio de Processamento de Sinais da UNICAMP (SPS)},
  year = {2010},
  pages = {35--37}
}
      


Pinto-Cáceres, S. and da S. Torres, R. Visualiza¸ cão para Sistemas de Recupera¸ cão de Imagens por Conteúdo 2010 I Simpósio de Processamento de Sinais da UNICAMP (SPS), pp. 49-51  inproceedings  
Abstract: Um sistema de Recupera¸ cão de Imagens por Conteúdo (CBIR) oferece mecanismos necessários para busca e recupera¸ cão de imagens baseando-se em características visuais como cor, textura, forma, etc. Em um processo de busca de imagens, a apresenta¸ cão de resultados é um componente essencial, pois representa a última etapa de um complexo processo e a obten¸ cão desses resultados é o motivo da existência do sistema. Consequentemente, a visualiza¸ cão dos resultados pode determinar o sucesso ou o fracasso de um sistema CBIR. Técnicas de visualiza¸ cão são valiosas ferramentas na explora¸ cão de grandes quantidades de dados, como cole¸ cões de imagens. Contudo, técnicas para visualizar imagens resultantes de sistemas CBIR têm sido pobremente exploradas.
Como parte deste trabalho, foi realizado um estudo comparativo de várias técnicas de visualiza¸ cão para sistemas CBIR. Além disso, foi proposta uma técnica original denominada Clustering Set, que tenta suprir algumas das limita¸ cões identificadas em métodos da literatura. Dentre as características da técnica proposta, destacam-se o enfoque baseado no centro e o uso de técnicas de minera¸ cão de dados para representar a similaridade intrínseca entre as imagens. Resultados experimentais mostram que o método proposto supera outras estratégias de visualiza¸ cão, considerando-se diversos critérios, como satisfa¸ cão de usuário, grau de entendimento, etc.

BibTeX:
@inproceedings{Pinto-Caceres2010SPS,
  author = {Sheila Pinto-Cáceres and Ricardo da S. Torres},
  title = {Visualiza¸ cão para Sistemas de Recupera¸ cão de Imagens por Conteúdo},
  booktitle = {I Simpósio de Processamento de Sinais da UNICAMP (SPS)},
  year = {2010},
  pages = {49--51}
}
      


dos Santos, J.A., da S. Torres, R. and ao , A.X.F. Classifica¸ cão Semi-automática de Regiões em Imagens de Sensoriamento Remoto Utilizando Realimenta¸ cão de Relevância 2010 I Simpósio de Processamento de Sinais da UNICAMP (SPS), pp. 41-43  inproceedings  
Abstract: O uso de imagens de sensoriamento remoto (ISRs) como fonte de informa¸ cão em aplica¸ cões voltadas para o agronegócio é bastante comum. Nessas aplica¸ cões, saber como é a ocupa¸ cão espacial é fundamental. Entretanto, reconhecer e diferenciar regiões de culturas agrícolas em ISRs ainda não é uma tarefa trivial. Embora existam métodos automáticos propostos para isso, os usuários preferem muitas vezes fazer o reconhecimento manualmente. Isso acontece porque tais métodos normalmente são feitos para resolver problemas específicos, ou quando são de propósito geral, não produzem resultados satisfatórios fazendo com que o usuário tenha que revisar os resultados manualmente. Esse trabalho apresenta um método semi-automático para vetorizar regiões em ISRs utilizando realimenta¸ cão de relevância baseada em programa¸ cão genética.

BibTeX:
@inproceedings{Santos2010SPS,
  author = {Jefersson A. dos Santos and Ricardo da S. Torres and Alexandre X. Falcão},
  title = {Classifica¸ cão Semi-automática de Regiões em Imagens de Sensoriamento Remoto Utilizando Realimenta¸ cão de Relevância},
  booktitle = {I Simpósio de Processamento de Sinais da UNICAMP (SPS)},
  year = {2010},
  pages = {41--43}
}
      


Murthy, U., Fox, E.A., Chen, Y., Hallerman, E., da S. Torres, R., Ramos, E.J. and Falcão, T.R.C. Species identification: fish images with CBIR and annotations 2009 Joint International Conference on Digital Libraries (JCDL), pp. 435-436  inproceedings DOI  
Abstract: Fish species identification is critical to the study of fish ecology and management of fisheries...

BibTeX:
@inproceedings{Murthy2009Poster-JCDL,
  author = {Uma Murthy and Edward A. Fox and Yinlin Chen and Eric Hallerman and Ricardo da S. Torres and Evandro J. Ramos and Tiago R. C. Falcão},
  title = {Species identification: fish images with CBIR and annotations},
  booktitle = {Joint International Conference on Digital Libraries (JCDL)},
  year = {2009},
  pages = {435--436},
  doi = {http://doi.org/10.1145/1555400.1555500}
}
      


Andaló, F.A., da S. Torres, R. and Falcão, A.X. Shape Descriptors based on Tensor Scale 2008 Workshop of Theses and Dissertations, XXI Brazilian Symposium on Computer Graphics and Image Processing  inproceedings  
Abstract: Tensor scale is a morphometric parameter that unifies the representation of local structure thickness, orientation, and anisotropy, which can be used in several computer vision and image processing tasks. We exploit this concept for binary images and propose two shape descriptors -- Tensor Scale Descriptor with Influence Zones and Tensor Scale Contour Saliences. It also introduces a robust method to compute tensor scale, using a graph-based approach -- the image foresting transform. Experimental results are provided, showing the effectiveness of the proposed methods, when compared to other relevant methods with regard to their use in content-based image retrieval tasks.

BibTeX:
@inproceedings{Andalo2008WTD-SIBGRAPI,
  author = {Fernanda A. Andaló and Ricardo da S. Torres and Alexandre X. Falcão},
  title = {Shape Descriptors based on Tensor Scale},
  booktitle = {Workshop of Theses and Dissertations, XXI Brazilian Symposium on Computer Graphics and Image Processing},
  year = {2008}
}
      


Kozievitch, N.P., Falcão, T.R.C. and da S. Torres, R. A .Net Implementation of a Content-Based Image Search Component 2008 Sessão de Demos, XXIII Brazilian Symposium on Databases  inproceedings  
Abstract: This work describes a .Net implementation of a content-based image search component, used for deployment and assessment of an image annotation and retrieval tool. It has been designed in a flexible and portable way, under the technology perspective, where different descriptors can be easily configured for different contexts.

BibTeX:
@inproceedings{Kozievitch2008Demo-SBBD,
  author = {Nádia P. Kozievitch and Tiago R. C. Falcão and Ricardo da S. Torres},
  title = {A .Net Implementation of a Content-Based Image Search Component},
  booktitle = {Sessão de Demos, XXIII Brazilian Symposium on Databases},
  year = {2008}
}
      


Murthy, U., da S. Torres, R., Fox, E.A., Venkatachalam, L., Yang, S. and Gonçalves, M.A. From concepts to implementation and visualization: tools from a team-based approach to ir 2008 31st Annual International ACM SIGIR Conference on Research and Development in Information Retrieval, pp. 889  inproceedings  
Abstract: Researchers have been studying and developing teaching materials for information retrieval (IR), such as [3]. Toolkits also have been built that provide hands-on experience to students. For example, IR-Toolbox [4] is an effort to close the gap between the student' understanding of IR concepts and real-life indexing and search systems. Such tools might be good for helping students in non-technical areas such as in the Library and Information Science field to develop their conceptual model of search engines. However, they do not cover emerging topics and skills, such as content-based image retrieval (CBIR) and fusion search. Although there is open source software (such as those in http://www.searchtools.com/tools/tools-opensource.html) that can be used to teach basic and advanced IR topics, they require a student to have high-level technical knowledge and to spend a long time to gain a practical understanding of these topics.

We present a new and rapid approach to teach basic and advanced IR topics, such as text retrieval, web-based IR, CBIR, and fusion search, to Computer Science (CS) graduate students. We designed projects that would help students grasp the abovementioned IR topics. Students, working in teams, were given a practical application to start with -- the Superimposed Application for Image Description and Retrieval [5]. SAIDR (earlier, SIERRA) allows users to associate parts of images with multimedia information such as text annotations. Also, users may retrieve information in one of two 2 ways: (1) Perform text-based retrieval on annotations; (2) Perform CBIR on images and parts of images that look like a query image (or part of a query image).

Each team was asked to build an enhancement for this application, involving text retrieval and/or CBIR, in three weeks time. The sub-projects are described in Table 1. The outcome of this activity was that students learned about IR concepts while being able to relate their applicability to a real world problem (Figure 1). Details of these projects may be found at http://collab.dlib.vt.edu/runwiki/wiki.pl?TabletPcImageRetrievalSuperimposedInformation. We will demonstrate the tools developed along with the IR concepts they illustrate (Table 1). We believe these tools may aid others to learn about basic and advanced topics in IR.



BibTeX:
@inproceedings{Murthy2008Poster-SIGIR,
  author = {Uma Murthy and Ricardo da S. Torres and Edward A. Fox and Logambigai Venkatachalam and Seungwon Yang and Marcos André Gonçalves},
  title = {From concepts to implementation and visualization: tools from a team-based approach to ir},
  booktitle = {31st Annual International ACM SIGIR Conference on Research and Development in Information Retrieval},
  year = {2008},
  pages = {889}
}
      


Macário, C., Senra, R., Medeiros, C.B., Lamparelli, R., Zullo, J., Rocha, J.V., Madeira, E.R.M., Martins, E., Baranauskas, M.C., Leite, N.J. and da S. Torres, R. Monitoramento de Safras via Web: Um Caso de Sucesso em Pesquisa Multidisciplinar 2007 Sexto Congresso Brasileiro de Agroinformática (SBIAgro)  inproceedings  
Abstract: Este trabalho descreve o projeto WebMaps em desenvolvimento na UNICAMP, um esfor¸ co multidisciplinar envolvendo ciências agrárias e de computa¸ cão, cujo objetivo é desenvolver uma plataforma baseada em servi¸ cos Web para o planejamento agro-ambiental. Requer pesquisa de ponta voltada à especifica¸ cão e à implementa¸ cão de software que combina dados de satélite, dados provenientes de sensores terrestres e dados de produ¸ cão agrícola. Visa fornecer subsídios para políticas públicas e tomada de decisões no planejamento agrícola.

BibTeX:
@inproceedings{Macario2007Poster-SBIAGRO,
  author = {Carla Macário and Rodrigo Senra and Claudia Bauzer Medeiros and Rubens Lamparelli and Jurandir Zullo and Jansle V. Rocha and Edmundo R. M. Madeira and Eliane Martins and Maria Cecília Baranauskas and Neucimar J. Leite and Ricardo da S. Torres},
  title = {Monitoramento de Safras via Web: Um Caso de Sucesso em Pesquisa Multidisciplinar},
  booktitle = {Sexto Congresso Brasileiro de Agroinformática (SBIAgro)},
  year = {2007}
}
      


Pedronette, D.C.G. and da S. Torres, R. RecS-DL: Uma plataforma para Servi¸ cos de Recomenda¸ cão em Bibliotecas Digitais 2007 Sessão de Demos, XXII Brazilian Symposium on Databases, pp. 15-20  inproceedings  
Abstract: Esta demonstra¸ cão visa ilustrar o funcionamento de uma plataforma de servi¸ cos de recomenda¸ cão recentemente criada no Instituto de Computa¸ cão da Unicamp. Baseada em Servi¸ cos Web e interfaces bem definidas, a ferramenta possibilita que engines de recomenda¸ cão, baseados em diferentes técnicas, possam ser desenvolvidos, instalados e testados no arcabou¸ co definido. A modelagem de dados realizada e a utiliza¸ cãoo de Servi¸ cos Web tornam a plataforma independente de domínio de aplica¸ cão e portável sob o ponto de vista tecnológico.

BibTeX:
@inproceedings{Pedronette2007Demo-SBBD,
  author = {Daniel C. G. Pedronette and Ricardo da S. Torres},
  title = {RecS-DL: Uma plataforma para Servi¸ cos de Recomenda¸ cão em Bibliotecas Digitais},
  booktitle = {Sessão de Demos, XXII Brazilian Symposium on Databases},
  year = {2007},
  pages = {15--20}
}
      


Pedronette, D. and da S. Torres, R. Uma plataforma de Serviços de Recomendação para Bibliotecas Digitais 2007 VI Workshop de Teses e Dissertações em Bancos de Dados, XXII Brazilian Symposium on Databases, pp. 51-56  inproceedings  
Abstract: Esta artigo apresenta um trabalho de mestrado que tem como foco de pesquisa servi¸ cos de recomenda¸ cão para bibliotecas digitais. A disserta¸ cão apresenta como contribui¸ cão uma plataforma de servi¸ cos de recomenda¸ cão baseada em servi¸ cos web e interfaces bem definidas, que possibilita que engines de recomenda¸ cão baseados em diferentes técnicas, possam ser desenvolvidos, instalados e testados no arcabou¸ co definido. A modelagem de dados realizada e a utiliza¸ cão de servi¸ cos web tornam a plataforma independente de domínio de aplica¸ cão e portável sob o ponto de vista tecnológico. Experimentos preliminares envolvendo bases de dados reais e bibliotecas digitais de diferentes domínios foram conduzidos para valida¸ cão da plataforma. Etapas seguintes do programa de mestrado consistem no estudo de arcabou¸ cos formais para descri¸ cão e documenta¸ cão da plataforma.

BibTeX:
@inproceedings{Pedronette2007WTDBD-SBBD,
  author = {D. Pedronette and R. da S. Torres},
  title = {Uma plataforma de Serviços de Recomendação para Bibliotecas Digitais},
  booktitle = {VI Workshop de Teses e Dissertações em Bancos de Dados, XXII Brazilian Symposium on Databases},
  year = {2007},
  pages = {51--56}
}
      


Montoya-Zegarra, J.A., Leite, N.J. and da S. Torres, R. Efficient and Effective Content-Based Image Retrieval Framework for Fingerprint Databases 2006 V Workshop of Theses and Dissertations, XXI Brazilian Symposium on Databases, pp. 77-82  inproceedings  
Abstract: Two kinds of fingerprint identification approaches have been proposed in the literature to reduce the number of one-to-many comparisons during fingerprint image retrieval, namely, exclusive and continuous classification. Although exclusive classification approaches reduce the number of comparisons, they present some shortcomings, including fingerprint ambiguous classification, and unbalanced fingerprint classification distribution. On the other side, continuous classification approaches have not been further studied. In this context, we propose an original continuous approach to guide the search and the retrieval in fingerprint image databases considering both effectiveness and retrieval speed. For that purposes, we use feature extraction and indexing methods considering the textural and directional information contained in fingerprint images. Preliminary results of our work involves a comparative study of several textural image descriptors obtained by combining different types of the Wavelet Transform with similarity measures. From our experiments we can conclude that the best retrieval accuracy was achieved by combining Gabor Wavelets (GWs) with the Square Chord similarity measure. Furthermore, the presence of noise and distortions in fingerprint images have affected the overall retrieval accuracy.

BibTeX:
@inproceedings{Montoya-Zegarra2006WTDBD-SBBD,
  author = {Javier A. Montoya-Zegarra and Neucimar J. Leite and Ricardo da S. Torres},
  title = {Efficient and Effective Content-Based Image Retrieval Framework for Fingerprint Databases},
  booktitle = {V Workshop of Theses and Dissertations, XXI Brazilian Symposium on Databases},
  year = {2006},
  pages = {77--82}
}
      


Montoya-Zegarra, J.A., Leite, N.J. and da S. Torres, R. Multiresolution Features for Fingerprint Image Retrieval 2006 Workshop of Theses and Dissertations, XIX Brazilian Symposium on Computer Graphics and Image Processing  inproceedings  
Abstract: This paper presents a real time system to guide the search and the retrieval in fingerprint image databases considering both retrieval accuracy and speed. For that purposes, we use multiresolution-based feature extraction and indexing methods considering the textural information inherent to fingerprint images. The extracted feature vectors are used to compute the distance between the fingerprint query image to all the fingerprints in the database and the N most similar images are then retrieved. The focus of this work is to study the utility of multiresolution transforms on the domain of fingerprint recognition.

BibTeX:
@inproceedings{Montoya-Zegarra2006WTD-SIBGRAPI,
  author = {Javier A. Montoya-Zegarra and Neucimar J. Leite and Ricardo da S. Torres},
  title = {Multiresolution Features for Fingerprint Image Retrieval},
  booktitle = {Workshop of Theses and Dissertations, XIX Brazilian Symposium on Computer Graphics and Image Processing},
  year = {2006}
}
      


Murthy, U., da S. Torres, R. and Fox, E.A. SIERRA - A Superimposed Application for Enhanced Image Description and Retrieval 2006 Research and Advanced Technology for Digital Libraries, 10th European Conference (ECDL), pp. 540-543  inproceedings DOI  
Abstract: In this demo proposal, we describe our prototype application, SIERRA, which combines text-based and content-based image retrieval and allows users to link together image content of varying document granularity with related data like annotations. To achieve this, we use the concept of superimposed information (SI), which enables users to (a) deal with information of varying granularity (sub-document to complete document), and (b) select or work with information elements at sub-document level while retaining the original context.

BibTeX:
@inproceedings{Murthy2006Demo-ECDL,
  author = {Uma Murthy and Ricardo da S. Torres and Edward A. Fox},
  title = {SIERRA - A Superimposed Application for Enhanced Image Description and Retrieval},
  booktitle = {Research and Advanced Technology for Digital Libraries, 10th European Conference (ECDL)},
  year = {2006},
  pages = {540--543},
  doi = {http://doi.org/10.1007/11863878%5C_63}
}
      


da S. Torres, R. Sistemas de Informa¸ cão para o Gerenciamento de Imagens: Aplica¸ cões e Desafios de Pesquisa 2006 Grandes Desafios da Sociedade Brasileira de Computa¸ cão  inproceedings  
Abstract: A cria¸ cão de novos dispositivos para obten¸ cão e armazenamento de imagens tem possibilitado a dissemina¸ cão de grandes cole¸ cões de imagens nas mais diversas aplica¸ cões: medicina, sensoriamento remoto, sistemas de informa¸ cão de biodiversidade, biometria, dentre outros. As solu¸ cões existentes para o gerenciamento integrado de imagens e dados espe¸ cíficos de cada aplica¸ cão baseiam-se na cria¸ cão de infra-estruturas para a armazenamento de dados e no desenvolvimento de ferramentas para realiza¸ cão de consultas a cada tipo de dado isoladamente. Por um lado, sistemas e ferramentas espe¸ cíficas têm sido propostos para a busca de imagens levando-se em conta suas descri¸ cões textuais associadas. Por outro lado, várias pesquisas vêm sendo realizadas visando a especifica¸ cão e a defini¸ cão de algoritmos que manipulam o conteúdo das imagens (objetos e suas propriedades de forma, cor e textura).

BibTeX:
@inproceedings{Torres2006GrandesDesafios-SBC,
  author = {Ricardo da S. Torres},
  title = {Sistemas de Informa¸ cão para o Gerenciamento de Imagens: Aplica¸ cões e Desafios de Pesquisa},
  booktitle = {Grandes Desafios da Sociedade Brasileira de Computa¸ cão},
  year = {2006}
}
      


Vemuri, N.S., da S. Torres, R., Shen, R., Gonçalves, M.A., Fan, W. and Fox, E.A. A Content-Based Image Retrieval Service for Archaeology Collections 2006 Research and Advanced Technology for Digital Libraries, 10th European Conference (ECDL), pp. 438-440  inproceedings DOI  
Abstract: Archeological sites have heterogeneous information ranging from different artifacts, image data, geo-spatial information, chronological data, and other relevant metadata. ETANA-DL, an archaeology digital library, provides various services by integrating the heterogeneous data available in different collections. This demonstration presents an initial prototype for searching DL objects based on the image content, using the Content-Based Image Search Component (CBISC) from Virginia Tech/State University of Campinas.

BibTeX:
@inproceedings{Vemuri2006Demo-ECDL,
  author = {Naga Srinivas Vemuri and Ricardo da S. Torres and Rao Shen and Marcos André Gonçalves and Weiguo Fan and Edward A. Fox},
  title = {A Content-Based Image Retrieval Service for Archaeology Collections},
  booktitle = {Research and Advanced Technology for Digital Libraries, 10th European Conference (ECDL)},
  year = {2006},
  pages = {438--440},
  doi = {http://doi.org/10.1007/11863878%5C_37}
}
      


da S. Torres, R., Falcão, A.X., Zhang, B., Fan, W., Fox, E.A., Gonçalves, M.A. and Calado, P. A new framework to combine descriptors for content-based image retrieval 2005 ACM International Conference on Information and Knowledge Management (CIKM), pp. 335-336  inproceedings DOI  
Abstract: In this paper, we propose a novel framework using Genetic Programming to combine image database descriptors for content-based image retrieval (CBIR). Our framework is validated through several experiments involving two image databases and specific domains, where the images are retrieved based on the shape of their objects.

BibTeX:
@inproceedings{Torres2005CIKM,
  author = {Ricardo da S. Torres and Alexandre X. Falcão and Baoping Zhang and Weiguo Fan and Edward A. Fox and Marcos André Gonçalves and Pável Calado},
  title = {A new framework to combine descriptors for content-based image retrieval},
  booktitle = {ACM International Conference on Information and Knowledge Management (CIKM)},
  year = {2005},
  pages = {335--336},
  doi = {http://doi.org/10.1145/1099554.1099654}
}
      


da S. Torres, R., Medeiros, C.B. and Falcão, A.X. Ambiente de Gerenciamento de Imagens e Dados Espaciais para Desenvolvimento de Aplicações em Biodiversidade 2005 XVIII Concurso de Teses e Dissertações, XXV Congresso da Sociedade Brasileira de Computação  inproceedings  
Abstract: There is a wide range of environmental applications requiring sophisticated management of several kinds of data, including spatial data and images of living beings. However, available information systems offer very limited support for managing such data in an integrated manner. This thesis provides a solution to combine these query requirements, which takes advantage of current digital library technology to manage networked collections of heterogeneous data in an integrated fashion. The research contributes to solve problems of specification and implementation of biodiversity information systems that manage images of species, textual descriptions and spatial data in an integrated way.

BibTeX:
@inproceedings{Torres2005CTD-SBC,
  author = {Ricardo da S. Torres and Claudia B. Medeiros and Alexandre Xavier Falcão},
  title = {Ambiente de Gerenciamento de Imagens e Dados Espaciais para Desenvolvimento de Aplicações em Biodiversidade},
  booktitle = {XVIII Concurso de Teses e Dissertações, XXV Congresso da Sociedade Brasileira de Computação},
  year = {2005}
}
      


da S. Torres, R., Medeiros, C.B., Gonçalves, M.A. and Fox, E.A. An OAI compliant content-based image search component 2004 ACM/IEEE Joint Conference on Digital Libraries (JCDL), pp. 418  inproceedings DOI  
Abstract: Advances in data storage and image acquisition technologies haveenabled the creation of large image datasets In order to deal withthese data, appropriate information systems (e.g. image digital libraries) have been developed to efficiently manage such collections. One of the most common retrieval approaches is to employ so--called Content--Based Image Retrieval (CBIR) systems. Basically, these systems try to retrieve images similar to auser--defined pattern (e.g. image example). Their goal is to support image retrieval based on content properties (e.g. shape, color, or texture), which are often encoded in terms of image descriptors This demonstration presents a new CBIR system based on configurable components. The main novelty resides in its Content--Based ImageSearch Component (CBISC) that supports queries on image collections CBISC is based on the OAI principles, and thus provides an easy--to--install search engine to support querying images by content As with the OAI protocol, queriesare posed via HTTP requests and the responses are encoded in XML CBISC encapsulates multidimensional index structures to speed up the search process. Furthermore, it supports the use of different image descriptors (metric and non--metric; color, texture, and shapedescriptors; with 1D or 2D feature vectors), which canbe easily combined to yield improved effectiveness. We will show that this search component can be tailored for particular image collectionsby a trained designer, who carries out a clearly defined set of pilot experiments to select the appropriate descriptors. Image descriptors are typically domain and usage--dependent Further, agiven image can be associated with very many descriptors However, standard CBIR methods only support a fixed set of descriptors CBISC, instead, allows progressive extension of the descriptor base Figure 1 presents a screen shotshowing the CBISC Configuration Tool developed to support CBISC designers in the configuration process Basically, this process concerns the description/definition of both the image descriptors thatwill be used to retrieve images by content, and the image database to which the CBISC is related The XML file generated in this process is used during CBISC execution.

BibTeX:
@inproceedings{Torres2004Demo-JCDL,
  author = {Ricardo da S. Torres and Claudia Bauzer Medeiros and Marcos André Gonçalves and Edward A. Fox},
  title = {An OAI compliant content-based image search component},
  booktitle = {ACM/IEEE Joint Conference on Digital Libraries (JCDL)},
  year = {2004},
  pages = {418},
  doi = {http://doi.org/10.1145/996350.996485}
}
      


da S. Torres, R., Medeiros, C.B., Dividino, R.Q., Figueiredo, M.A., Gonçalves, M.A., Fox, E.A. and Richardson, R. Using digital library components for biodiversity systems 2004 ACM/IEEE Joint Conference on Digital Libraries (JCDL), pp. 408  inproceedings DOI  
Abstract: Biodiversity information systems (BISs) involve all kinds of heterogeneous data, which include ecological and geographical features However, available information systems offer very limited support for managing such data in an integrated fashion, and such integration is often based on geographic coordinates alone. Furthermore, such systems do not fully support image content management (e.g. photos of landscapes or living organisms), a requirement of many BIS end--users. In order to meet their needs, these users -- e.g. biologists, environmental experts -- often have to alternate between distinct biodiversity and image information systems to combine information extracted from them. This cumbersome operational procedure is forced on users by lack of interoperability among these systems. This hampers the addition of new data sources, as well as cooperation among scientists. The approach provided in this project to meet these issues is based on taking advantage of advances in Digital Library (DL) innovations to integrate networked collections of heterogeneous data. It focuses on creating the basis for a biodiversity information system under the digital library perspective, combining new techniques of content--based image retrieval and database query processing mechanisms. This approach solves the problem of system switching, and provides users with a flexible platform from which to tailor a BIS to their needs. The main contributions of this project are the following: (a) a generic architecture for managing heterogeneous collections, based ondigital library components, to access heterogeneous biodiversity data sources (text and images), that allows combining text--based andcontent--based queries in a seamless way; and (b) a new component, for content--based image search, integrated into that architecture. The proposed architecture has been implemented by using DL components which are mostly new or recently developed. Furthemore, its implementation uses the Open Archives Initiative (OAI) protocol as a basis for interoperability. This architecture is easily extensible, and provides users a considerable degree of flexibility in data management. To illustrate our claim that this architecture can be applied to several domains, we are investigating its application in building a biodiversity information system on fish species. This solution solves many current problems in this kind of system, allowing handling of images and textual information in an integrated fashion. A new Content-Based Image Search Component has been developed to support queries on image collections. Since this component is based onthe OAI principles, it provides an easy-to-install search engine toquery images by content. It can be readily tailored for a particular collection by a trained designer, who carries out a clearly definedset of pilot experiments. It supports the use of different image descriptors, which can be chosen from the pilot experiment, and theneasily combined to yield improved effectiveness. In addition, it encapsulates a multidimensional index structure to speed up the search process, that also can be easily configured for different image collections.

BibTeX:
@inproceedings{Torres2004Poster-JCDL,
  author = {Ricardo da S. Torres and Claudia Bauzer Medeiros and Renata Queiroz Dividino and Mauricio Augusto Figueiredo and Marcos André Gonçalves and Edward A. Fox and Ryan Richardson},
  title = {Using digital library components for biodiversity systems},
  booktitle = {ACM/IEEE Joint Conference on Digital Libraries (JCDL)},
  year = {2004},
  pages = {408},
  doi = {http://doi.org/10.1145/996350.996474}
}