Special session at INNS Big Data “Soft computing methods in large-scale content-based image retrieval”

at the INNS-BigData’ 2015, San Francisco, USA, 8-10 August 2015 (http://innsbigdata.org/).

Up to now the only way to search vast collections of images and video which are generated every day in tremendous amount is by keywords and meta tags or by just browsing them. Emergence of content-based image retrieval (CBIR) in 1990s enabled automatic retrieval of images to a certain extent. The goal can be searching for similar images to the query image, classification of the query image or retrieving images of a certain class. Such content-based image matching remains challenging problem of computer science as large multimedia datasets generate big data to analyze. Image matching consists in two relatively difficult tasks: identifying objects on images and fast searching through large collections of identified objects. Identifying objects on images is still a challenge as the same objects and scenes can be viewed under different imaging conditions. There are many previous work dedicated to such formed problem. Some of them bases on color representation, textures, shape or edge detectors. Recently local invariant features has gained wide popularity. To find similar images to a query one, we need to compare all local features descriptors of all images usually by some distance measure. Such a comparison is enormously time consuming and there is ongoing research to speed up the process. Yet the current state-of-the-art in case of high dimensional, computer vision applications is not fully satisfactory. The literature presents countless methods and variants. Mostly they are based on some form of approximate search. One of the solutions of the problem can be hashing descriptor vectors. The session will focus on application of different soft computing techniques in content-based image retrieval and classification, such us:

• fuzzy image annotation,
• fuzzy similarity measures for visual feature matching,
• soft computing methods to visual data cleaning,
• soft computing methods for fast feature comparing,
• novel soft computing algorithms and structures for CBIR,
• soft computing methods for visual objects and image classification,
• evolutionary algorithms for CBIR,
• and other topics on large scale CBIR.

Paper Submission

Please submit you paper by March 22, 2015.

• Prospective authors should submit a full-length draft manuscript (8 pages), including figures, tables and references using the Procedia Computer Science Standard template (available here) without page numbers.
• The full papers should be submitted in PDF format to the INNS-BigData’2015 EasyChair online submission website: https://easychair.org/conferences/?conf=innsbigdata2015, please remember to choose the track: Special Session on Soft computing methods in large-scale content-based image retrieval.
• Conference content will be submitted for inclusion into Procedia Computer Science and indexed by Scopus.
• All papers submitted to INNS-BigData’2015 will be checked for plagiarism including self-plagiarism. If a paper is found to fall in the category of plagiarism, the paper will be automatically rejected. A list of Authors of the papers identified in the category of plagiarism will be forwarded to Elsevier.
• After the Conference, a selected number of Best Papers will be expanded and revised for possible inclusion in Special Issues of Big Data Research journal.

Organizers:

Marcin Korytkowski
Czestochowa University of Technology, Poland
e-mail: marcin.korytkowski@iisi.pcz.pl

Leszek Rutkowski
Czestochowa University of Technology, Poland
http://www.kik.pcz.czest.pl/~rutkowski/
e-mail: leszek.rutkowski@iisi.pcz.pl

Rafal Scherer
Czestochowa University of Technology, Poland
http://iisi.pcz.pl/Rafal_Scherer/
e-mail: rafal.scherer@iisi.pcz.pl

We look forward to meeting you in San Francisco!