Conditional models for contextual human motion recognition

• Published in: Tenth IEEE International Conference on Computer Vision (ICCV'05) Volume 1 Vol. 2; pp. 1808 - 1815 Vol. 2
• Main Authors: Sminchisescu, C., Kanaujia, A., Zhiguo Li, Metaxas, D.
• Format: Conference Proceeding
• Language: English
• Published: IEEE 2005
• Subjects: Artificial intelligence; Character generation; Chromium; conditional models; discriminative models; Feature extraction; feature selection; Gold; Hidden Markov models; human motion recognition; Humans; Image recognition; Inference algorithms; Markov random fields; multiclass logistic regression; Optical filters; optimization
• ISBN: 076952334X; 9780769523347
• ISSN: 1550-5499; 2380-7504
• DOI: 10.1109/ICCV.2005.59

We present algorithms for recognizing human motion in monocular video sequences, based on discriminative conditional random field (CRF) and maximum entropy Markov models (MEMM). Existing approaches to this problem typically use generative (joint) structures like the hidden Markov model (HMM). Therefore they have to make simplifying, often unrealistic assumptions on the conditional independence of observations given the motion class labels and cannot accommodate overlapping features or long term contextual dependencies in the observation sequence. In contrast, conditional models like the CRFs seamlessly represent contextual dependencies, support efficient, exact inference using dynamic programming, and their parameters can be trained using convex optimization. We introduce conditional graphical models as complementary tools for human motion recognition and present an extensive set of experiments that show how these typically outperform HMMs in classifying not only diverse human activities like walking, jumping. running, picking or dancing, but also for discriminating among subtle motion styles like normal walk and wander walk