2nd International Workshop on Interactive Multimedia Retrieval

held at the ACM Multimedia Conference, October 2022, Lisbon, Portugal

About

With the growth of multimedia content in many areas of our daily lives, the ability to efficiently retrieve relevant content becomes increasingly vital for the content to be relevant. Despite the recent advances in deep learning-based semantic extraction and annotation methods, in many cases, a single query is insufficient to express an information need precisely enough to produce the desired result. Therefore, iterative and more importantly interactive retrieval methods need to be employed. There are however few dedicated venues for work in the area of interactive retrieval and even fewer that consider more than one media domain. This workshop aims at addressing this lack by creating a venue for the discussion of the current state of the art in interactive multimedia retrieval, independent of the media type.

The 2nd International Workshop on Interactive Multimedia Retrieval will be held in conjunction with the ACM Multimedia Conference 2022 in Lisbon.

Important Dates

  • Workshop Paper Submission: 1st 10th of July 2022

  • Notification: 29th of July 2022

  • Camera-ready Version: 7th of August 2022

  • Workshop: 14th of October 2022

Author Instructions

Despite the main conference using pre-recorded video presentations for the hybrid participants, the workshop will only have live presentations, either on-site or via Zoom. The Zoom link will be distributed via the Whova platform of the conference. Authors should prepare presentations of ca. 20 minutes in length. Authors are also expected to have a working demonstration of their interactive retrieval approaches for workshop participants to interact with during the Demo Session. Posters can be optionally used to accompany a demo but this is not required.

Schedule (tentative)

09:00 - 09:15 Welcome Message from the Chairs
09:15 - 09:40 Presentation 1: Combining Semantic and Visual Image Graphs for Efficient Search and Exploration of Large Dynamic Image Collections
09:40 - 10:05 Presentation 2: NeoCube: Graph-Based Implementation of the M3 Data Model
10:05 - 10:30 Presentation 3: Influence of Late Fusion of High-Level Features on User Relevance Feedback for Videos
10:30 - 11:00 Coffee Break
11:10 - 11:25 Presentation 4: Impact of Blind Image Quality Assessment on the Retrieval of Lifelog Images
11:25 - 11:50 Presentation 5: An Asynchronous Scheme for the Distributed Evaluation of Interactive Multimedia Retrieval
11:50 - 12:15 Presentation 6: VILT: Video Instructions Linking for Complex Tasks
12:15 - 14:00 Lunch Break / Demo Setup
14:00 - 14:30 Demo Flash Introductions
14:30 - 16:00 Demo Session
16:00 - 17:00 Panel Discussion & Closing

Organizers

  • Werner Bailer (Joanneum Research)

  • Jakub Lokoč (Charles University)

  • Luca Rossetto (University of Zurich)

  • Klaus Schoeffmann (Klagenfurt University)

Program Committee

  • George Awad (NIST)

  • Ralph Gasser (University of Basel)

  • Silvan Heller (University of Basel)

  • Oana Inel (University of Zurich)

  • Björn Þór Jónsson (IT University of Copenhagen)

  • Mathias Lux (Alpen-Adria-Universität Klagenfurt)

  • Shin'ichi Satoh (NII Japan)

  • Loris Sauter (University of Basel)

  • Florian Spiess (University of Basel)

  • Stefanos Vrochidis (Centre for Research and Technology Hellas)