REGIMVID at TRECVID2010: Semantic Indexing.
Nizar Elleuch, Mohamed Zarka, Issam Feki, Anis Ben Ammar, Adel M. Alimi · 2010
��Regim_4: The indexing process is based on the visual modality analysis and relationships within LSCOM Ontology to improve the detection of large set of semantic concepts. The visual modality analysis is orientated towards an automatic categorization of video contents to create relevance relationships between low-level descriptions and semantic contents according to a user point of view. ��Regim_5: This indexing system is based on Multimodal fuzzy fusion using positive rules extracted from LSCOM Ontology. The fusion process employs both a deduction and abduction reasoning engines. ��Regim_6: This indexing system is based on Multimodal fuzzy fusion using positive and negative rules extracted from LSCOM Ontology. The fusion process employs also both a deduction and abduction reasoning engines. In this paper, we describe an overview of a software platform that has been developed within REGIMVid project for TRECVID 2010 video retrieval experiments. The REGIMVID team participated in Semantic Indexing task. In TRECVID 2010, we explore several novel techniques to perform the detection of semantic concepts, including multi classifiers with supervised learning process, discriminative feature representation based on local keypoints, visual concept fusion using LSCOM rules, and also multimodal concept fusion using LSCOM Ontology.