Smart MMIR is subject to research and development. The Generic Multimedia Analysis Framework (GMAF) as well as the KM-EP system developed at the University of Hagen and its corresponding technologies, are the basis for Smart Multimedia Information Retrieval.
The Generic Multimedia Analysis Framework (GMAF), the Multimedia Feature Graphs (MMFG) as a representational data structure for MMIR features, and the Graph Code concept for fast and effective indexing and retrieval, are core technologies for various projects in the area of research and development. Therefore, on this page, a summary of the current state of these projects is given.
Where to start
- SMAP Conf.’20: Towards AI based MMIR
- BIRDS Conf.’21: Query Construction’
- SMAP-Jrnl.’21: MMIR with Graph Codes
- ICIVR Conf.’21: Graph Codes Projections
- MDPI-BigData.’21: Effective MMIR
- SMAP-Jrnl.’21: Towards explainability of MMIR
GMAF Source, Docs and Tools
- Installation Manual (FP2021)
- Java Doc Folder
- Merge Detection XML
- Term Normalizer
- Sample Evaluation
- Sample Annotations
- XML Schema for Data Exports
- Sample XML Export
- Plugin Development (Video)
Smart Multimedia Information Retrieval
- MDPI Jrnl.’21: Fast and Effective Retrieval f. Big Data
- ArgStrength ’21: Sem. Feature Detection and Named Entities
- BIRDS Conf.’21: Query Construction and Result Presentation
- ICIVR Conf.’21: Graph Codes Projections
- MDPI Jrnl.’20: AI based MMIR on Smartphones
- SMAP Conf.’20: Towards AI based MMIR
- HILOD21 – a High-Level-Of-Detail Dataset
- MDPI Jrnl. ’21: Explainable MMFGs
- TREC2021 News Track
- Bibtex File