Intelligent Information Analysis and Unstructured Information Mining

We are investigating fundamental research issues involved in the automatic extraction of intelligent information from unstructured information such as images, audio, video, rich text, as well as heterogeneous distributed sensor networks. The common goal across all projects in this area is to apply machine learning, signal processing and database techniques to extract information that can be used for accessing the unstructured data at a semantic level. This analysis will help build solutions for semantic enrichment and repurposing of the unstructured data.

Active Projects

  ·  Content Adaptation and MPEG 21   The transcoding system adapts video, images, audio and text to the individual pervasive devices using a new framework that allows the content to be summarized , translated and converted, on-the-fly.
  ·  CueVideo   CueVideo is an ongoing research project to address the challenges that arise in the creation, indexing and use of large video databases.
  ·  MARVEL  MARVEL is a MPEG-7 Video Search Engine that helps organize the large and growing amounts of multimedia data (e.g., video, images, audio) by using machine learning techniques to automatically label its content.
  ·  SLAM  Semantic Learning and Analysis of Multimedia project investigates the application of machine learning algorithms that provide generic trainable procedures to model the semantic concepts