Vasileios Mezaris
Electrical and Computer Engineer, Ph.D.
homepage curriculum vitae projects research demos downloads publications contact j b c p

Dr. Vasileios Mezaris is a Senior Researcher with the Information Technologies Institute / Centre for Research and Technology Hellas, Thessaloniki, Greece, since June 2007, and since October 2013 he is a Senior Researcher Grade B. His research interests include image and video analysis, event detection in multimedia, machine learning for multimedia analysis, content-based and semantic image and video retrieval, applications of image and video analysis in specific domains (TV broadcasting and News, medical images, ecological data, educational and cultural applications).

Dr. Mezaris has co-authored more than 30 papers in refereed journals, 10 book chapters, 130 papers in international conferences, and 3 patents. He serves as Associate Editor for the IEEE Signal Processing Letters, served as Associate Editor for the IEEE Transactions on Multimedia (2012-2015), and serves regularly as a reviewer for many international journals and conferences. He has participated in many research projects, currently being the Coordinator of the EC H2020 research projects InVID (2016-2018) and MOVING (2016-2019). He is a Senior Member of the IEEE and a member of the Technical Chamber of Greece.

(Mar. 2017) New data release: we released concept detection scores for the IACC.3 dataset used in the TRECVID AVS Task from 2016 and on (600 hr of internet archive videos). See our downloads page for details and links.
(Jan. 2017) A new and exciting H2020 Innovation Action, EMMA, just started!
(Oct. 2016) Our special Issue on "Multimedia in Ecology" has been published in Multimedia Systems (Springer): vol. 22, no. 6, pp. 709-782, Nov. 2016.
(Oct. 2016) We successfully participated in the Ad-hoc Video Search (AVS) and Multimedia Event Detection (MED) tasks of TRECVID 2016. In AVS, our best fully-automatic run was ranked 2nd-best (MXinfAP=0.051, compared to 0.054 and 0.040 for the 1st and 3rd best-performing teams using fully-automatic systems). In MED we also achieved very good results, with our best run reaching MInfAP@200=0.475.
(Sept. 2016) Our special Issue on "Fine-grained categorization in ecological multimedia" has been published in Pattern Recognition Letters (Elsevier): vol. 81, pp. 51-117, Oct. 2016.
(July 2016) Call for Papers, Special Issue on Deep Learning for Mobile Multimedia, in the ACM Transactions on Multimedia Computing, Communication and Applications (TOMM). Paper submission deadline: Oct. 31, 2016.
(Feb. 2016) Accelerated Kernel Subclass Discriminant Analysis (AKSDA) and SVM combination: An efficient GPU-accelerated dimensionality reduction and classification method, for very high-dimensional data. Achieve SoA classification results, consistently higher than Kernel SVM approaches, at orders-of-magnitude shorter training times. More info and software download here.
(Jan. 2016) Two new and exciting H2020 projects, InVID and MOVING, just started or are about to start!
(Sept. 2015) New interactive on-line video analysis service: you upload videos via a web interface, and it performs shot/scene segmentation and visual concept detection (several times faster than real-time; uses our new concept detection engine). Results are displayed in an interactive user interface, which allows navigating through the video structure (shots, scenes), viewing the concept detection results for each shot, and searching by concepts within the video. Try this service now!
(Sept. 2015) Updated machine-to-machine video analysis service: a REST service receiving a video's URL, performing temporal segmentation to shots and scenes as well as visual concept detection, and returning the results in a machine-processable MPEG-7-compliant XML file. Updated in Sept. 2015 with our new concept detection engine. A short video-demo: You can try this service (email us, mentioning your name and university / company affiliation, to receive a link and instructions).
(May 2015) Call for Papers, 2nd ACM International Workshop on Human-centered Event Understanding from Multimedia (HuEvent15) at the 23rd ACM Multimedia Conference (ACM MM 2015), Brisbane, Australia, 26-30 October 2015. Submission deadline: June 30, 2015.
(May 2015) The program and keynote of the Human Memory-Inspired Multimedia Organization and Preservation (HMMP15) Workshop @ IEEE ICME 2015, Torino, Italy, July 3, 2015, is announced here.
(Feb. 2015) Call for Papers, Human Memory-Inspired Multimedia Organization and Preservation (HMMP15) Workshop @ IEEE ICME 2015, Torino, Italy, June 29 - July 3, 2015. Submission deadline: March 30, 2015.
(Feb. 2015) Call for Papers, Special Issue on "Fine-grained Categorization in Ecological Multimedia", Pattern Recognition Letters, Elsevier. Submission deadline: March 15, 2015.
(Nov. 2014) A new, very fast machine learning method for big data problems was developed. This is a generic learning method, applicable to a wide range of problems that involve learning from big data. Tested in large-scale multimedia event detection problems, it produces better results than both Kernel- and Linear-SVMs, while its training is one or two orders of magnitude faster than that of Linear-SVMs. Check out our TRECVID presentation on this.
(Nov. 2014) Tutorials on "Video Hyperlinking" and related video analysis technologies were delivered at the IEEE Int. Conf. on Image Processing (ICIP'14), Paris, France, Oct. 2014, and at ACM Multimedia (MM'14), Orlando, FL, USA, Nov. 2014. The tutorials' slides are available in slideshare, in three parts:
- Part A: Motivation and Vision.
- Part B: Video Fragment Creation and Annotation for Hyperlinking.
- Part C: Insights into Hyperlinking Video Content.
(Sept. 2014) New software (v1.4 of CERTH's real-time video shot and scene segmentation software) was publicly released in September. Check out the downloads page for details and links.
(July 2014) New dataset (CERTH Image Blur Dataset) was publicly released during the last couple of weeks. Check out the downloads page for details and links.
(May 2014) The Proceedings of the 1st International Workshop on Social Events in Web Multimedia @ ACM International Conference on Multimedia Retrieval (ICMR 2014) (Glasgow, UK, April 1-4, 2014) are available here.
(Apr. 2014) Call for Papers, Human-centered Event Understanding from Multimedia (HuEvent14) @ ACM Multimedia 2014, Orlando, FL, USA, November 3-7, 2014. Submission deadline: June 30, 2014.
(Apr. 2014) Call for Papers, 3rd ACM Intl. Regular & Data Challenge Workshop on Multimedia Analysis for Ecological Data (MAED 2014) @ ACM Multimedia 2014, Orlando, FL, USA, November 3-7, 2014. Submission deadline: June 30, 2014. The Challenge dataset is available.
(Apr. 2014) Call for Multimedia Grand Challenge Solutions: VideoLectures.NET Challenge (MediaMixer, transLectures): Temporal segmentation and annotation of lecture videos @ ACM Multimedia 2014, Orlando, FL, USA, November 3-7, 2014. Submission deadline: June 29, 2014. The Challenge dataset is already available (ask for access details).
(Apr. 2014) Call for Participation, Social Event Detection (SED) and Synchronization of Event Media (SEM) benchmarking tasks at MediaEval 2014. Early registration (by May 1st, 2014) is encouraged!
(Jan. 2014) Best Paper Award at the 20th International Conference on MultiMedia Modeling (MMM'14, Dublin, Ireland) for the paper "A Comparative Study on the Use of Multi-Label Classification Techniques for Concept-Based Video Indexing and Annotation", authored by Fotini Markatopoulou, Vasileios Mezaris and Ioannis Kompatsiaris.
(Jan. 2014) Tutorial on Re-using Media On The Web at the 23rd International World Wide Web Conference (WWW'14), Seoul, Korea, April 7-11, 2014.

© 2015 Vasileios Mezaris