iwemsa 2012 - The International Workshop on Advances in Large-Scale Multimedia Data Collection, Mining and Retrieval

Recent years have witnessed an explosive proliferation of digital multimedia data due to high processor speeds, the wide availability of massive storage devices, high-resolution video cameras, fast networks, and the advent of cloud computing. The huge and ever growing volume of multimedia data in online or in-house repositories, such as Flickr and YouTube, have led to a much higher requirement for multimedia data mining and retrieval in terms of effectiveness and efficiency. However, existing approaches and systems cannot meet such requirements. The increasing demand for mining and retrieval over large-scale multimedia data has attracted a wealth of research efforts in multimedia community. On the other hand, to facilitate research in this area, it is necessary to systematically evaluate and compare the performance of various approaches and systems. Thus sufficiently large benchmark datasets are expected. Recently, more and more research groups are individually constructing such datasets. There is a need to unify these individual efforts into the creation of a unified large-scale repository which would benefit the entire multimedia community.
This workshop aims to bring together researchers and industrial practitioners interested in large-scale multimedia data collection, mining and retrieval. The workshop will provide a venue for the participants to discuss key research issues on large-scale multimedia data. It will collect and seek the recent important research works in large-scale multimedia data collection, mining and retrieval, summarize the available resources in this area, and exploit potential challenges and possible advanced solutions in terms of theory and practice. Topics of interests include, but are not limited to, the following:
Construction, unification and evolution of large-scale multimedia data collections (including images, video, audio and other multi-modal data)
Feature extraction and representation for large-scale multimedia data
Efficient and scalable technologies for multimedia storage
Indexing, clustering, retrieval and browsing for large-scale multimedia data collections
Large scale object discovery
Compact visual representation for fast extraction and matching
Bag-of-Words model and extension for large scale object and duplication retrieval
Interactive image search
Real-time face/object search in videos
Image search applications to image editing
Parallelization of indexing and clustering algorithms
Collaborative multimedia tagging for distributed users
Semantic annotation, object recognition, copy detection, near/partial-duplicate detection on large-scale multimedia data
Scalable and distributed machine learning and data mining technologies for multimedia
Interfaces for exploring, browsing and visualizing large-scale multimedia data collections
Interactive search and retrieval for large-scale multimedia data collections
Machine learning for large-scale multimedia applications
Scalable and distributed systems for large-scale multimedia analysis and management
Exploration of social networks for large-scale multimedia computing