SISAP 2026: 19th International Conference on Similarity Search and Applications Faculty of Informatics, Masaryk University Brno, Czechia, October 14-16, 2026

19th International Conference on Similarity Search and Applications, SISAP 2026

October 14–16, 2026, Brno, Czech Republic

CORE Rank B conference, https://www.sisap.org/2026/

Topics in brief: Learned Similarity, Embeddings, Vector Databases, Scalable Retrieval, Multimedia and Multimodality, Similarity Models and Theory, Demos and Applications

Scope

The International Conference on Similarity Search and Applications (SISAP) focuses on research in similarity-based data management and retrieval, with emphasis on embedding-based methods, vector databases, and machine-learning-driven similarity search.

SISAP covers similarity models, indexing and query processing, scalable and distributed similarity systems, learned and adaptive techniques, and similarity-aware database architectures supporting high-dimensional and multimodal data.

Originating from metric indexing research, SISAP is the only international conference dedicated exclusively to similarity search, spanning theory, systems, evaluation, and applications across data management, information retrieval, and machine learning.

Topics of Interest

The SISAP conference solicits original research contributions on similarity search and its applications. Topics of interest include, but are not limited to:

Similarity Models and Theory

• Models of similarity and dissimilarity in metric and non-metric spaces
• Intrinsic dimensionality, concentration phenomena, hubness, and discriminability
• Manifolds, embeddings, and geometric properties of similarity spaces
• Theoretical foundations and limits of similarity search and indexing

Learning and Representations

• Feature extraction and representation learning for similarity search
• Metric learning and learned similarity measures
• Embeddings from self-supervised and foundation models
• Multimodal and cross-modal similarity representations

Similarity Queries and Processing

• Similarity queries and operators (k-NN, range, reverse NN, top-k, diversity queries)
• Exact, approximate, and probabilistic similarity search
• Similarity joins, ranking, filtering, and aggregation
• Query semantics and languages for similarity-based data
• Cross-modal similarity search

Indexing and Scalable Systems

• Indexing and access methods for similarity search
• Graph-based, tree-based, hashing, quantization, and hybrid approaches
• Learned and adaptive index structures
• Parallel, distributed, and GPU-accelerated similarity processing
• Dynamic, streaming, and update-aware similarity systems

Similarity-Aware Data Management

• Similarity search in database and data management systems
• Vector databases and similarity-native storage engines
• Query optimization and execution for similarity workloads
• Integration of similarity search with relational, graph, and hybrid systems
• Cloud-native and large-scale similarity services

Evaluation and Benchmarks

• Evaluation methodologies and cost models for similarity processing
• Benchmark datasets, workloads, and experimental frameworks
• Accuracy–efficiency trade-offs and reproducibility

Applications

• Similarity search in multimedia, scientific, industrial, and emerging data domains
• Similarity search in healthcare, sports, robotics, security, and other fields
• Dense retrieval and semantic search
• Recommendation systems and personalization
• Search and question-answering within content collections

Invited Speakers

Josef Sivic - Czech Institute of Informatics, Robotics, and Cybernetics. Czech Technical University in Prague, Czech Republic

Bio: Josef Sivic is a distinguished researcher at the Czech Institute of Informatics, Robotics and Cybernetics (CIIRC) at the Czech Technical University in Prague, where he leads the Intelligent Machine Perception group and the ELLIS Unit Prague. He is currently on leave from a senior researcher position at Inria Paris, where he remains an external collaborator with the Willow team. Josef received his PhD from the University of Oxford in 2006 under the supervision of Professor Andrew Zisserman, followed by a postdoctoral appointment at MIT’s CSAIL with Professor William Freeman, and obtained his habilitation degree from École Normale Supérieure in Paris in 2014. His research lies at the intersection of computer vision, machine learning, and robotics, with a particular focus on learning visual representations from large-scale, weakly supervised, and multimodal data. He has made influential contributions to visual recognition, image and video retrieval, semantic segmentation, visual localization, geometric matching, and 3D understanding, as well as to learning from videos, narrated data, and demonstrations for embodied and robotic tasks. His work has been published in top venues such as CVPR, ICCV, ECCV, NeurIPS, and TPAMI. He has an outstanding publication record, with an h-index of 82 and roughly 60,000 citations.

Allan Hanbury - TU Wien, Complexity Science Hub Vienna, Austria

Bio: Allan Hanbury is Professor for Data Intelligence, head of the Data Science Research Unit, and Faculty Representative (responsible for financial affairs and internationalisation) at the Faculty of Informatics, TU Wien, Austria. He is also faculty member of the Complexity Science Hub Vienna. He was scientific coordinator of the EU-funded Khresmoi Project on medical and health information search and analysis, and is co-founder of contextflow, the spin-off company commercialising the radiology image search technology developed in the Khresmoi project. He is coordinator of DoSSIER, a Marie Curie Innovative Training Network, educating 15 doctoral students on domain-specific systems for information extraction and retrieval. He also coordinated the EU-funded VISCERAL project on evaluation of algorithms on big data, and the EU-funded KConnect project on technology for analysing medical text. He is author or co-author of over 180 publications in refereed journals and refereed international conferences. He contributes to research and innovation strategy development in Austria and Europe, and regularly gives talks on topics related to his research. He has an outstanding publication record, with an h-index of 53 and roughly 16,000 citations.

Regular Papers

Full papers (from 9 to 14 pages in Springer LNCS format) are expected to be descriptions of complete technical work, whereas short papers (of up to 8 pages) can describe innovative approaches or preliminary results which may nevertheless require more work to mature. Vision papers and other position papers should be submitted as short research papers. Page limits include references. Any appendices, if needed, can only be posted online, and the reviewers are not expected to take them into account. 

Demonstration Papers

Demonstration papers (of up to 8 pages in Springer LNCS format) should provide the motivation for the demonstrated concepts, the information about the technology and the system to be demonstrated, and should state the significance of the contribution. The scenarios within which the demonstrated system applies should also be explained. Evaluation criteria for the demonstration proposals include: novelty, technical advances and challenges, and the overall practical attractiveness of the demonstrated system. A demonstration submission consists of a paper and an additional 1-page appendix (in PDF format) that illustrates how the demo will be conducted on-site at SISAP. This additional content will not be published in the conference proceedings, should the submission be accepted.

Doctoral Symposium Papers

A submission to the doctoral symposium consists of a paper and an additional 1-page appendix (both in PDF format), which must be single-author and written by the student alone. The paper should be no longer than 6 pages in Springer LNCS format (plus up to 2 pages of references). The paper must describe the problem being addressed, an outline of the planned methodology, contributions made so far, and the work lying ahead as part of the author’s PhD study. The additional 1-page appendix will not be published in the conference proceedings, should the submission be accepted. This appendix should describe the benefits that would be obtained by attending the doctoral symposium, namely the student’s motivation to attend SISAP, and their advisor’s word on how the student would benefit by attending the Doctoral Symposium.

SISAP Indexing Challenge

SISAP Indexing Challenge is an event for researchers and practitioners aimed at advancing the state-of-the-art in large-scale similarity data management. The challenge provides a platform to showcase innovative solutions and push the boundaries of efficiency and effectiveness in indexing, filtering, and searching. The results provide valuable comparisons of competing approaches and their implementations from given viewpoints and environments. It is expected that participants prepare a detailed report of their solution and results in a typical SISAP's short-paper format, which will be included in the LNCS proceedings of SISAP 2026.

Learn more: https://www.sisap.org/2026/indexingchallenge.html