The fall conference season is beginning in earnest, and Kitware’s Computer Vision team, led by Dr. Anthony Hoogs, will be actively participating and presenting cutting-edge work at a variety of events. With paper acceptance rates often no higher than 25%, and sometimes even more competitive, the number of papers the Kitware team is presenting is a testament to their high-quality and innovative research. Further, in addition to presenting work, Kitware team members are actively contributing to the vision community by serving as co-chairs on organizing committees.
The Military Symposium on Big Data is a gathering of leading government and industry experts, with discussion focused on the future of big data R&D and how that can be leveraged for defense and homeland security. As part of this year’s “High-Value Applications for Defense and Intelligence – Emerging Lessons Learned” session, Dr. Arslan Basharat, who is stepping in for Dr. Anthony Hoogs, will give an invited talk on large-scale video analytics. Dr. Hoogs has previously presented at this event, as it is an effective forum bringing together interesting, invited speakers from government and industry.
GEOINT is the largest public event for the intelligence community, with more than 3,000 attendees and invited talks by the directors of most of the US intelligence agencies. Kitware has participated at various levels for the past several years, and this year will contribute to the lightning talks on Monday, October 13, as part of GEOINT FOREWORD. Anthony Hoogs will present a short talk, “Large-Scale Human Geography from Crowd-Sourced Multimedia.”
ACM Multimedia brings together experts from a diverse set of fields to focus on the state-of-the-art in multimedia technology. Kitware, in collaboration with Simon Fraser University (SFU), authored the paper “Segmental Multi-way Local Pooling for Video Recognition,” which will be presented as part of the poser session on October 22. The paper was authored by Ilseo Kim, Sangmin Oh, Arash Vahdat (SFU), Kevin Cannons (SFU), Amitha Perera, and Greg Mori (SFU).
IEEE ICCV is the leading global computer vision conference, held this year in Sydney, Australia. Dr. Anthony Hoogs is on the organizing committee as the Corporate Relations Co-Chair with responsibilities for raising funding from industry, and Kitware is a Silver Sponsor for the event. Additionally, stemming from our leading-edge research and development programs, our Computer Vision team authored two papers that were accepted for this year’s event.
The first, “Pyramid Coding for Functional Scene Element Recognition in Video Scenes,” was written by Eran Swears, Anthony Hoogs, and Kim Boyer (RPI). In collaboration with Rensselaer Polytechnic Institute through Eran Swear’s Ph.D. thesis, the paper is on recognizing objects in video scenes efficiently by the activity happening around them. This paper was sponsored by a DARPA STTR.
The second paper, “Compositional Models for Video Event Detection: A Multiple Kernel Learning Latent Variable Approach,” was authored by Arash Vahdat (Simon Fraser University), Kevin Cannons (SFU), Greg Mori (SFU), Sangmin Oh, and Ilseo Kim. In collaboration with Simon Fraser University, the paper is on recognizing complex events in internet videos, and was sponsored by Kitware’s ALADDIN project.
More information on our Computer Vision work is available on our website or in our Computer Vision brochure. If you would like to meet with Kitware team members at any of these events to discuss collaboration or employment opportunities, please contact us at Kitware@kitware.com.