Informace o projektu
Centrum pro multi-modální interpretaci dat velkého rozsahu
- Kód projektu
- GBP103/12/G084 (kod CEP: GBP103/12/G084)
- Období řešení
- 1/2012 - 12/2018
- Investor / Programový rámec / typ projektu
-
Grantová agentura ČR
- Projekty na podporu excelence v základním výzkumu
- Fakulta / Pracoviště MU
-
Fakulta informatiky
- prof. Ing. Pavel Zezula, CSc.
- RNDr. Michal Batko, Ph.D.
- doc. RNDr. Vlastislav Dohnal, Ph.D.
- RNDr. David Novák, Ph.D.
- Spolupracující organizace
-
Univerzita Karlova
- Odpovědná osoba RNDr. Pavel Pecina, Ph.D.
- Odpovědná osoba prof. Ing. Josef Psutka, CSc.
- Odpovědná osoba prof. Ing. Jiří Matas, Ph.D.
The project aims at exploiting large collections of unlabeled multi-modal data, mainly video footage, to further state-of-the-art in video, audio and natural language understanding, interpretation, annotation and retrieval by combining unsupervised and semi-supervised learning. It will address problems that are very difficult (some probably impossible) to solve in a single modality by adopting an interdisciplinary approach. Progress in individual research areas - vision, natural language processing and speech will be achieved by co-training and by exploiting results of other modalities as cross-training data. For efficient processing of large data collections appearing in all modalities, the project will also concern generic problems of organization, indexing, and searching based on similarity that is fundamental for building real-life applications.
Publikace
Počet publikací: 49
2014
-
DISA at ImageCLEF 2014 Revised: Search-based Image Annotation with DeCAF Features
Rok: 2014, druh: Účelové publikace
-
DISA at ImageCLEF 2014: The search-based solution for scalable image annotation
CLEF 2014 Evaluation Labs and Workshop, Online Working Notes, rok: 2014
-
Performance Study of Independent Anchor Spaces for Similarity Searching
The Computer Journal, rok: 2014, ročník: 57, vydání: 11, DOI
-
Person Tagging in Still Images by Fusing Face and Full-body Detections
International Conference on Advances in Multimedia (MMEDIA 2014), rok: 2014
-
Privacy-preserving Outsourced Similarity Search
Journal of Database Management, rok: 2014, ročník: 25, vydání: 3, DOI
-
Rank Aggregation of Candidate Sets for Efficient Similarity Search
25th International Conference on Database and Expert Systems Applications (DEXA 2014 ), rok: 2014
-
Semantically Consistent Human Motion Segmentation
Proceedings of 25th International Conference on Database and Expert Systems Applications (DEXA 2014), rok: 2014
-
Towards Fast Multimedia Feature Extraction: Hadoop or Storm
Proceedings of 2014 IEEE International Symposium on Multimedia (ISM), rok: 2014
2013
-
A Key-Pose Similarity Algorithm for Motion Data Retrieval
Proceedings of 12th International Conference on Advanced Concepts for Intelligent Vision Systems (ACIVS 2013), LNCS 8192, rok: 2013
-
Content-based annotation and classification framework: a general multi-purpose approach
Proceedings of the 17th International Database Engineering & Applications Symposium, rok: 2013