Hive: Scalable Transport of Vision Data

Many computer vision systems require transport of images and distribution of processing tasks. Various systems exist for general purpose transport, however these did not support computer vision directly so we have created a small framework for computer vision distribution and transport. Our framework is called Hive and is available for academic use. The main project page for Hive is at the HCT web site. The video below demonstrates a quality-of-view analysis system built on Hive to automatically switch to the view containing the most faces.

Publications and Activities

Vision System Development Through Separation of Management and Processing, Amir Afrah, Gregor Miller and Sidney Fels, Workshop on Multimedia Information Processing and Retrieval (MIPR), San Diego, California, U.S.A., December 2009
  PDF   BibTeX
.

Rapid Vision Application Development using Hive, Gregor Miller, Amir Afrah and Sidney Fels, International Conference on Computer Vision Theory and Applications (VISAPP), Lisboa, Portugal, February 2009
  PDF   BibTeX
.

Hive: A Distributed System for Vision Processing, Amir Afrah, Gregor Miller, Donovan Parks, Matthias Finke and Sidney Fels, International Conference on Distributed Smart Cameras (ICDSC), Palo Alto, California, U.S.A., September 2008
  PDF   BibTeX
.

Videos
Automatic Camera Selection based on Faces