Home
vitrivr is an open source full-stack content-based multimedia retrieval system originating in the Databases and Information Systems research group as a culmination of various research project and dissertations. Its modular architecture makes it a powerful tool for both, research and people which have to work with large multimedia collections. vitrivr follows a three-tier system architecture with
- a frontend (User Interface)
- a retrieval engine This is the wiki for the retrieval engine!
- a database
as seen in the following architecture diagram.
Source: https://doi.org/10.1145/3643489.3661120
This is the documentation for vitrivr-engine, the retrieval engine.
We have split the documentation into several parts which we think will help you unerstand how to setup and use vitrivr-engine.
- Getting Started - For those who do not care and just want to follow a guide to setup and use vitrivr-engine.
- Documentation - In-depth documentation of how vitrivr-engine works.
- Example - A more advanced example.
- Development - A guide for developers who either want to contribute towards vitrivr-engine or want to built their own project, using vitrivr-engine.
We kindly ask you to refer to one the following papers in publications mentioning or employing vitrivr-engine.
Expansion with Boolean retrieval:
Loris Sauter, Ralph Gasser, Laura Rettig, Heiko Schuldt, and Luca Rossetto. 2024. General Purpose Multimedia Retrieval with vitrivr at LSC'24. In Proceedings of the 7th Annual ACM Workshop on the Lifelog Search Challenge (LSC '24). Association for Computing Machinery, New York, NY, USA, 47–52. https://doi.org/10.1145/3643489.3661120
Introduction of vitrivr-engine:
Gasser, R., Arnold, R., Faber, F., Schuldt, H., Waltenspül, R., & Rossetto, L. (2024, January). A New Retrieval Engine for Vitrivr. In International Conference on Multimedia Modeling (pp. 324-331). Cham: Springer Nature Switzerland. https://link.springer.com/chapter/10.1007/978-3-031-53302-0_28