Keywords to Keyframes: Video Analytics for Archival Research

July 15, 2013, 13:00 | Workshop, Ubuntu, Multicultural Center

The target audience for this workshop consists of scholars with research interests related to the way that visual media impacts culture.

Description + Schedule

This workshop will serve scholars of any level of technical expertise who are interested in studying images as part of their work in the digital humanities using a hybrid method that combines machine analytics (keyframes) and crowd-sourced tagging (keywords). Facilitating a discussion and training session featuring up to twenty participants, we will demo the Large Scale Video Analytics (LSVA) workbench for moving and still image analysis and archiving. The LSVA is a web portal developed through a collaboration among the National Center for Supercomputing Applications (NCSA) and the Extreme Science and Engineering Discovery Environment (XSEDE), the IML (Institute for Multimedia Literacy), and ICHASS (the Institute for Computing in the Humanities, Arts and Social Science). The LSVA has customized the prominent Medici content management system, a multimedia database which has served scholars worldwide. The LSVA requires no software installation, though we do require online access. Further, we will provide access to IM2Learn, a free software package for image analysis developed by the NCSA. There will be no CFP associated with this workshop; rather, we would like participants to self select.

The LSVA deploys the application of various algorithms for image recognition and visualization into the workflow that allows real-time analysis of video, as well as crowd-sourced content labeling such that the system becomes more valuable the more it is used.

In addition, the LSVA team has created and customized visualization tools that enhance research in several ways: novel visualizations employ spatial and temporal simultaneity, revealing unique aspects of a single film sequence; comparative visualizations represent relationships among multiple films within an archive; and, finally, the integration of visualization imagery becomes an input tag and a front end process that feeds the Medici content management system and enhances word-based labels, helping to close the semantic gap that occurs when words are applied to images.

  • 8:00-8:30 am: Introduction to basic concepts of computer vision and image retrieval
  • 8:30-9:00 am: Overview of standard research methodologies and those the LSVA extends
  • 9:00-9:15 am:BREAK
  • 9:15-10:00 am: Demo of LSVA system and walkthrough of interpreting output and requerying
  • 10:00-11:00 am: Hands-on with LSVA system: uploading, sorting, and analyzing moving and still images

Workshop Leader Bios

BIO: Virginia Kuhn is associate director of the Institute for Multimedia Literacy, an organized research unit in the School of Cinematic Arts at the University of Southern California. She was the 2009 recipient of the USC provost’s Award for Teaching with Technology, she co-chairs the Scholarly Interest Group on Media Literacy and Pedagogical Outreach for the Society for Cinema and Media Studies, and she serves on the editorial boards of several journals of media and technology. She joined USC in 2005 after successfully defending one of the first born-digital dissertations in the US, challenging archiving and copyright conventions. Committed to helping shape open source tools for scholarship, she published the first article created in Scalar, which appeared in the International Journal of Learning and Media and titled “Filmic Texts and the Rise of the Fifth Estate. She also serves on the editorial boards of several journals of media and technology.

BIO: Michael Simeone is the Associate Director for Research and Interdisciplinary Studies at ICHASS at the University of Illinois at Urbana-Champaign. Aside from his work on projects that engage the computational study of video, image-analysis of Great Lakes area historical maps, and the significance of social network analytics for the humanities, his research focuses on the intersection of humanities research procedures and data science. He received his PhD in English from the University of Illinois at Urbana-Champaign.