Staff R&D Engineer
Dr. Swears joined Kitware, Inc. as one of the first Research & Development Engineers in the Computer vision group in 2007. Since then, he has worked as the principle researcher or project lead on several DARPA and AFRL efforts. These projects aligned well with his interests, which include: pattern recognition, machine learning, graphical models, and in particular activity modeling and detection. Dr. Swears has performed research in motion pattern learning, scene decomposition, anomaly detection, tracking, event and activity detection, as well as video and chat text fusion using surveillance video, ground based sensors, and chat text of video.
Prior to working at Kitware, Dr. Swears joined the Discrimination Group at Lockheed Martin in Moorestown, NJ where he designed and developed algorithms for the Aegis system inter-continental ballistic missile tracker. His research included the development of multi-object tracking, track linking, genealogy, and event detection using solid state radar detections. After progressing to the Tracker Team Lead in 2005 he joined the GE Global Research Center as a contractor in 2006. While at GE Dr. Swears researched and developed algorithms for motion pattern learning and anomaly detection with computer vision applications.
Dr. Swears received his Ph.D. from Rensselaer Polytechnic Institute in 2015, an M.S. from Drexel University in Philadelphia PA in 2005, and a B.S. from Rensselaer Polytechnic Institute in 2001. The majority of his graduate studies were performed while working full-time at Lockheed Martin and Kitware. Dr. Swears has numerous publications that include papers in top tier conferences and journals.
- An End-to-End System for Content-Based Video Retrieval using Behavior, Actions, and Appearance with Interactive Query Refinement," Aug. 2015. , "
- Complex Activity Recognition using Granger Constrained DBN (GCDBN) in Sports and Surveillance Video," Jun. 2014. , "
- Pyramid Coding for Functional Scene Element Recognition in Video Scenes," Jun. 2014. , "
- Probabilistic sub-graph matching (pgram) for video and text fusion," MSS National Symposium on Sensor and Data Fusion (NSSDF), Oct. 2014. , "
- Pyramid Coding for Functional Scene Element Recognition in Video Scenes," Dec. 2013. , "
- Recognizing Activity Based Scene Elements in Video," Jun. 2013. , "
- Modeling temporal interactions with interval temporal bayesian networks for complex activity recognition," IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 35, no. 0162-8828, p. 2468–2483, Oct. 2013. , "
- Learning and Recognizing Complex Multi-Agent Activities with Applications to American Football Plays," Jan. 2012. , "
- Automatic activity profile generation from detected functional regions for video scene analysis," Video Analytics for Business Intelligence, Studies in Computational Intelligence, vol. 409, p. 241–269, Jan. 2012. , "
- Complex activity recognition using granger constrained dynamic bayesian network," Learning Workshop, Apr. 2011. , "
- A Large-scale Benchmark Dataset for Event Recognition in Surveillance Video," p. 3153–3160, Jun. 2011. , "
- Functional Scene Element Recognition for Video Scene Analysis," Dec. 2009. , "
- Learning Motion Patterns in Surveillance Video using HMM Clustering," Jan. 2008. , "