Abstract

We address the problem of predicting the physical state of a an agent performing a known activity. In particular we are interested in predicting human movement during complex composite activities. Our proposed framework combines a graphical model that extends the Sequential Interval Network (SIN) [2] for modeling global temporal structure of activities with a low level dynamic system for modeling the dynamics of the physical state. Specifically, two sets of new hidden state variables are added: one with respect to the temporal structure and one with respect to time. A mapping factor is defined to ensure these variables values remain consistent and hence allows fusing the two sources of information. We then derive an inference algorithm for computing the posterior densities of the hidden variables. The system can run in an on-line predictive mode to recognize on-going activity and make predictions arbitrarily far in the future during execution of the activity. Experiments illustrate that the long term prediction performance benefits from the knowledge about the temporal structure of the activity while short term prediction performance is improved by incorporating the dynamics of physical state.

Download

• Messy code
• Dataset

Papers

[1] Augmenting Physical State Prediction Through Structured Activity Inference. ICRA 2015

[2] From Stochastic Grammar to Bayes Network:Probabilistic Parsing of Complex Activity, CVPR 2014