5th Workshop Visual Information Coding Meets Machine Learning: Large-Scale Challenges (VICML'18)
Poznań, Poland, 9 - 12 September, 2018
The aim of the Workshop is to present new solutions addressing challenges of large-scale multimedia applications. The Workshop will focus on new groundbreaking technologies exploring large collections of images, videos, audio and text documents. Besides the recent impressive achievements in the domain of local descriptors developed for various types of multimedia contents and deep learning neural networks, most of these technologies face numerous shortcomings related to the complexity, memory and robustness. Also, most popular relational database and data warehouses management systems do not provide efficient solutions for handling visual information. These problems have a multidisciplinary nature and require expertise from computer vision, image processing, coding, machine learning and database management. Therefore, this Workshop should serve as an opportunity for multidisciplinary researchers to exchange their experience and ideas on the above open problems.
Topics
The Workshop will cover but it is not limited to:
- New methods of efficient content representation targeting complexity-memory-robustness trade-off
- New techniques of data encoding for fast retrieval and recognition
- New methods of benchmarking and optimization targeting comparison of description based and neural network based systems
- Machine learning architectures for visual recognition
- Deep learning algorithms and models
- Convolutional networks
- Mobile visual search systems
- Fine-grained recognition systems
- Cloud computing and architectures for computer vision
- Security and privacy-preserving retrieval and recognition systems
- Novel algorithms based on computational intelligence for analyzing online analytical processing (OLAP) data
- Computer network anomaly and intrusion detection
- Optimization of processes associated with relational databases performance
- SQL language extensions for handling multimedia, large text objects, and spatial data
- NoSQL databases for visual data
- Machine learning applications in road safety, astronomy or medicine
- Applications
Paper submission
- Authors should submit draft papers (as Postscript, PDF or MSWord file).
- The total length of a paper should not exceed 10 pages IEEE style (including tables, figures and references). IEEE style templates are available here.
- Papers will be refereed and accepted on the basis of their scientific merit and relevance to the workshop.
- Preprints containing accepted papers will be published on a USB memory stick provided to the FedCSIS participants.
- Only papers presented at the conference will be published in Conference Proceedings and submitted for inclusion in the IEEE Xplore® database.
- Conference proceedings will be published in a volume with ISBN, ISSN and DOI numbers and posted at the conference WWW site.
- Conference proceedings will be indexed in BazEkon and submitted for indexation in: Thomson Reuters - Conference Proceedings Citation Index, SciVerse Scopus, Inspec, Index Copernicus, DBLP Computer Science Bibliography and Google Scholar
- Extended versions of selected papers presented during the conference will be published as Special Issue(s).
- Organizers reserve right to move accepted papers between FedCSIS events.