Content-based query of image databases: inspirations from text retrieval
Pattern Recognition Letters - Selected papers from the 11th scandinavian conference on image analysis
Hierarchical classification using a frequency-based weighting and simple visual features
Pattern Recognition Letters
The MedGIFT group at ImageCLEF 2008
CLEF'08 Proceedings of the 9th Cross-language evaluation forum conference on Evaluating systems for multilingual and multimodal information access
Overview of the CLEF 2009 medical image retrieval track
CLEF'09 Proceedings of the 10th international conference on Cross-language evaluation forum: multimedia experiments
CLEF'09 Proceedings of the 10th international conference on Cross-language evaluation forum: multimedia experiments
The CLEF 2005 cross–language image retrieval track
CLEF'05 Proceedings of the 6th international conference on Cross-Language Evalution Forum: accessing Multilingual Information Repositories
The CLEF 2004 cross-language image retrieval track
CLEF'04 Proceedings of the 5th conference on Cross-Language Evaluation Forum: multilingual Information Access for Text, Speech and Images
| Hi-index | 0.00 |
MedGIFT is a medical imaging group of the Geneva University Hospitals and the University of Geneva, Switzerland. Since 2004, the group has participated ImageCLEF each year, focusing on the medical imaging tasks. For the medical image retrieval task, two existing retrieval engines were used: the GNU Image Finding Tool (GIFT) for visual retrieval and Apache Lucene for text. Various strategies were applied to improve the retrieval performance. In total, 16 runs were submitted, 10 for the image-based topics and 6 for the case-based topics. The base-line GIFT setup used for the past three years obtained the best results among all our submissions. For medical image annotation two approaches were tested. One approach is using GIFT for retrieval and kNN (k-Nearest Neighbors) for classification. The second approach used the Scale-Invariant Feature Transform (SIFT) with a Support VectorMachine (SVM) classifier. Three runs were submitted, two with the GIFT-kNN approach and one using the common results of the two approaches. The GIFT-kNN approach gave stable results. The SIFT-SVM approach did not achieve the expected performance, most likely due to the SVM Kernel used that was not optimized.