Probabilistic Modelling, Reasoning, and Decision-Making

  Active Information Fusion for Decision Making

Details...

An Approximate Algorithm for Value-of-Information Computation in Influence Diagrams

A common scenario in decision making is that, given a decision problem on hand, and a lot of information sources, what information are deserved to be used, or used first? Value-of-information analysis provides a straightforward means for selecting the best next observation to make, for determining whether it is better to gather additional information or to act immediately. However, it requires a consideration of the value of making all possible sequences of observations, which is intractable in real applications. Thus people have to make a myopic assumption: only one additional test will be performed, even when there is an opportunity to make a large number of observations. However, such an assumption is not reasonable in a lot of applications. Thus, we propose an approximate algorithm to compute non-myopic value-of-information efficiently. With this algorithm, the value-of-information can be computed efficiently for any group of information sources, thus makes it possible for people to choose best information sources for decision-making.

  An Efficient Inference Algorithm for Bayesian Networks slides

A Bayesian Network (BN) is a directed acyclic graph (DAG) that represents a joint probability over a set of variables. BN is a very powerful probabilistic knowledge representation and reasoning tool for partial beliefs under uncertainty. It has been widely applied in many applications, e.g., Artificial Intelligence, Medical diagnosis, etc. In a Bayesian Network, a probabilistic inference is the procedure of computing the posterior probability of query variables given a collection of evidences. However, when a BN is large, the inference is difficult and time-consuming. In this study, we propose an algorithm that efficiently carries out the inferences whose query variables and evidence variables are restricted to a subset of the set of the variables in a BN. The algorithm successfully combines the advantages of two popular inference algorithms - variable elimination and clique tree propagation.

Human Computer Interaction

 A Decision-Theoretic Framework for Emotion Recognition and User Assistance slides

In this study, we develop a general unified dynamic decision-theoretic framework based on Influence Diagrams for simultaneously modeling both affective states recognition (stress, frustration, fatigue, etc) and user assistance for human-computer interaction systems. Affective state recognition is achieved through active probabilistic inferences from the available sensory data of multiple-modality sources. User assistance is automatically achieved by balancing the benefits of keeping user in productive affective states and the costs of performing user assistances. We focus on some theoretical issues within the framework: user affect recognition, active sensory action selection, and user assistance. In addition, we try to use both synthetic data and real data to validate the framework. We built a non-invasive real world system to recognize user stress and fatigue from four-modality evidences, namely physical appearance features, physiological measures, user performance and behavioral data. The experiment results are promising.

Computer Vision

 Robust Visual Tracking Using Case-Based Reasoning with Confidence

This research proposes a simple but robust framework for visual object tracking in a video sequence. Compared with the existing tracking techniques, our proposed tracking technique has two significant contributions. First, a Case-Based Reasoning (CBR) paradigm is introduced to track the non-rigid object robustly under significant appearance changes without drifting away. Second, it can provide an accurate confidence measurement for each tracked object so that the tracking failures can be identified successfully. Specifically, under this framework, the appearance changes of the object being tracked can be adapted dynamically during tracking via an adaption mechanism of CBR. Hence, an accurate 2D tracking model can be maintained online for each image frame during tracking. Therefore, the proposed tracking technique possesses a self-recovery capability so that the object can be tracked robustly under significant appearance changes without error accumulation. Application was focused on the development of a real-time face tracking system. Via the proposed framework, the built real-time face tracker can track the human face robustly at 26 frames per second under various face orientations, significant facial expression and external illumination changes. Such a system can be easily generalized to track other objects by only updating its case base.

Demos

 Automatic Facial Action Unit Recognition

A system that could automatically analyze the facial actions in real time has applications in a wide range of different fields. However, developing such a system is always challenging due to the richness, ambiguity, and the dynamic nature of facial actions. Although a number of research groups attempt to recognize facial action units (AUs) by either improving facial feature extraction techniques, or the AU classification techniques, these methods often recognize AUs or certain AU combinations individually and statically, ignoring the semantic relations among AUs and the dynamics of AUs. Hence, these approaches cannot always recognize AUs reliably, robustly, and consistently. We propose a novel approach that systematically accounts for the relationships among AUs and their temporal evolutions for AU recognition. Specifically, we use a dynamic Bayesian network (DBN) to model the relationships among different AUs. The DBN provides a coherent and unified hierarchical probabilistic framework to represent probabilistic relations among various AUs and to account for the temporal changes in facial action development. Under our system, robust computer vision techniques are used to obtain AU measurements. And such AU measurements are then applied as evidences into the DBN for inferencing various AUs. The experiments show that the integration of AU relationships and AU dynamics with AU measurements yields significant improvement in AU recognition, especially for spontaneous facial expressions and under more realistic environment including illumination variation, face pose variation, and occlusion.

Demos


Publications:

Journal Papers:

Wenhui Liao, Weihong Zhang, Zhiwei Zhu, Qiang Ji, and Wayne Gray, "Toward a Decision-Theoretic Framework for Affect Recognition and User Assistance",  International Journal of Human-Computer Studies, vol.64, no.9, pp.847-873, 2006.

 Yan Tong, Wenhui Liao, and Qiang Ji, "Facial Action Unit Recognition by Exploiting Their Dynamic and Semantic Relationships", accepted by the IEEE Transactions on Pattern Analysis and Machine Intelligence (PAMI).

 Wenhui Liao and Qiang Ji, "Efficient Non-myopic Value-of-information Computation for Influence Diagrams", under submission

Conference Papers:

Wenhui Liao, Yan Tong, Zhiwei Zhu, and Qiang Ji, “Robust Face Tracking with a Case-base Updating Strategy”, to appear atthe Twentieth International Joint Conference on Artificial Intelligence (IJCAI-07), January 2007 (acceptance rate: 15.7%).

Wenhui Liao and Qiang Ji, “Efficient Active Fusion for Decision-making via VOI Approximation”, the Twenty-First National Conference on Artificial Intelligence (AAAI-06) (presentation), Boston, July 2006.  (acceptance rate: 22.1%).

 Wenhui Liao, “Dynamic and Active Information Fusion for Decision Making under Uncertainty”, The Twenty-First National Conference on Artificial Intelligence (AAAI) Doctoral Consortium, July 2006.

Zhiwei Zhu, Wenhui Liao, and Qiang Ji, "Robust Visual Tracking Using Case-based Reasoning with Confidence", the International Conference on Computer Vision and Pattern Recognition (CVPR'06), June, 2006.  (acceptance rate: 28.1%).

Yan Tong, Wenhui Liao, and Qiang Ji, "Inferring Facial Action Units with Causal Relations”, the International Conference on Computer Vision and Pattern Recognition (CVPR'06), June, 2006.  (acceptance rate: 28.1%).

Wenhui Liao, Weihong Zhang, Zhiwei Zhu, and Qiang Ji, "A Decision Theoretic Model for Stress Recognition and User Assistance", The Twentieth National Conference on Artificial Intelligence (AAAI-05) (presentation), pp. 529-534, 2005.   (acceptance rate: 18.4%).

Wenhui Liao, Weihong Zhang, Zhiwei Zhu, and Qiang Ji, "A Real-Time Human Stress Monitoring System Using Dynamic Bayesian Network", IEEE Workshop on Vision for Human-Computer Interaction (V4HCI), in conjunction with IEEE International Conference on Computer Vision and Pattern Recognition (CVPR'05), 2005.

Markus Guhe, Wenhui Liao, Zhiwei Zhu, Qiang Ji, Wayne D. Gray, & Michael J. Schoelles, "Non-intrusive measurement of workload in real-time", Human Factors and Ergonomics Society 49th Annual Meeting, 2005

Wenhui Liao, Weihong Zhang, and Qiang Ji, "A Factor Tree Inference Algorithm for Bayesian Networks and its Application", the 16th IEEE International Conference on Tools with Artificial Intelligence (ICTAI 2004), Boca Raton, FL, 2004 (presentation).

Shu-Ching Chen, Mei-Ling Shyu, Wenhui Liao, and Chengcui Zhang, "Scene Change Detection By Audio and Video Clues", Proceedings of the IEEE International Conference on Multimedia and Expo (ICME2002),August, 2002, Lausanne, Switzerland (presentation).

Dissertation and Thesis

Wenhui Liao, "Dynamic and Active Information Fusion for Decision Making under Uncertainty", Ph.D. dissertation proposal, Rensselaer Polytechnic Institute, May 2006.

 Wenhui Liao, "Video Scene Detection and Content-based Audio Classification", Master's thesis, Florida International University, May 2003.

Wenhui Liao, "Design and Implement a Search Engine for Medical Database", Bachelor's thesis, Xi'an Jiaotong University, July 2001.

Useful Links:

Bayesian Networks

Introductions:
Bayesian Network without tears--by Eugene Charniak
An Introduction to Bayesian Network Theory and Usage --by Todd A.Stephenson Charniak

Books:
Probabilisitic Reasoning in Intelligent Systems by Judea Pearl
Bayesian Networks and Decision Graphs by Finn V.Jensen
Probabilistic Networks & Expert Systems by R. G. Cowell, A. P. Dawid, S. L. Lauritzen, and D. J. Spiegelhalter

Research People:
David Heckerman
Steffen L. Lauritzen
Daphne Koller
Ross D. Shachter
more people...

Software
Software for Manipulating Belief Networks
Bayesian Network Repository
Murphy's BN Toolbox

Research Associations

AUAI: Association for Uncertaintly in Artificial Intelligence
AAAI: American Association for Artificial Intelligence
Decision Analysis Society
Human Factor and Ergonomics Society
Computer Vision List

Useful Digital Library

JSTOR
UAI Proceedings Search
CiteSeer: Scientific Literature Digital Library
Elsevier ScienceDirect

Related Conferences:

AAAI: National Conference on Artificial Intelligence
UAI:Conference on Uncertainty in Artificial Intelligence
IJCAI: International Joint Conference on Artificial Intelligence
NIPS: Neural Information Processing Systems Conference
ICTAI: IEEE International Conference on Tools with Artificial Intelligence
More AI conference ...

ICCV: IEEE International Conference on Computer Vision
CVPR: IEEE International Conference on Computer Vision and Pattern Recognition
ICPR: IEEE International Conference on Pattern Recognition

CDC: IEEE International Conference on Decision and Control

Fusion: IEEE International Conference on Information Fusion

CHI: Conference on Human Factors in Computing Systems
HFES: Human Factors and Ergonomics Society Annual Meeting

Computer Science Conference Rankings (For Reference Only)
Computer Science Journal Rankings (For Reference Only)