Center for Intelligent Information Retrieval

Multi-media Indexing and Retrieval (MIR)
Our Approach People in the group Papers (.gz postscript files)
Multi-media Indexing and Retrieval is a group within CIIR that concentrates on:
  1. techniques for locating and indexing text in images and scanned documents.
  2. techniques for directly indexing and retrieving images by content.
Current work includes:
  1. Extracting Text from images: A project to extract text from images.
  2. Word Spotting: A project to index handwritten archives by matching on word images.
  3. Image Retrieval by Appearance: A project to retrieve images from a database based on their visual similarity to a query image.
  4. Image Retrieval by Color: A project to retrieve advertisements from a database based on their color similarity to a query. You can see a demo .

The MIR groups work is supported by a number of sources including a grant from the US Patent and Trademarks Office and DARPA , a NSF Stimulate grant and a NSF CISE Research Infrastructure grant.
Back to Main CIIR