MLMD 2013 - Special Session on Machine Learning with Multimedia Data

Vast amounts of new data in a large variety of formats and media modalities are made available worldwide on a daily basis. Much of this information is not easily reachable; for instance, accessing relevant rich multimedia information (e.g. audio, images, video) is an extremely challenging problem. Machine learning is used with varying success for diverse tasks dealing with multimedia data. Many problems remain, including effective data representation schemes, semantic-enabled feature representations, algorithms capable of dealing with high-dimensional spatio-temporal data, fusion of multi-modal content features or techniques for enabling cross-modal access to information (e.g. textual queries of video recordings).
This session aims at bringing together researchers working on applying machine learning to multimedia data, including music, video, speech, or images. We welcome papers describing work in progress and encourage submissions that make datasets available to the community.
Topics covered by this special session include, but are not limited, to the following:
Feature extraction from multimedia data
Semantic content analysis, classification & annotation
Semi-supervised learning in multimedia data analysis
Learning techniques for cross-media enabled scenarios
Learning from user generated content in Social Media
Multimedia personalization and recommender systems
Multimedia machine learning in biometrics
Deep Learning applied to multimedia content
Optimal Learning from multimodal features
Semantic-aware interfaces for multimedia and cross-media navigation
Multimedia information retrieval
Spoken document retrieval
Speech recognition
Speaker recognition & diarization