CAP6938-02: Special Topics in Plan, Activity, and Intent Recognition
Schedule: Fall 2007, TR 1330-1445
Location: CL1 212
Professor: Dr. Gita Sukthankar
Plan recognition is the process of making inferences about other agents
based on observations, prior knowledge, and closed-world assumptions.
This synergistic area of AI research combines techniques from
human-computer interaction, autonomous and multi-agent systems, natural
language understanding, machine learning, and computer vision.
Objectives:
After completing the course, students should:
- understand current research issues in plan/activity/intent
recognition;
- learn the following major approaches and models: symbolic, probabilistic,
graphical, and game-theoretic;
- implement a software application that uses plan/activity/intent
recognition.
Additionally, students will refine their research, writing, and
presentation skills.
Assignments
Outline
- Introduction; Robocup Application Domain
- Elder-Care Application Domain; Event Hierarchy Circumscription
- Event Tracking in SOAR; Plan Recognition for Adversarial Domains
- Efficiency Improvements for Symbolic Plan Recognition
- Probabilistic Grammars
- Graphical Models
- Dynamic Bayes Networks (DBNs)
- Applications of Hidden Markov Models
- Review
- Exam Problems; Presentations; Questionnaire
- Partial Observable Markov Decision Processes (POMDPs)
- Model Learning for POMDPs; Emotion Recognition
- Plan Recognition in Spoken Dialogue Corpora; Unsupervised Learning of Object Categories
- Human Motion Prediction; Tactical Game Play
- Undirected Graphical Models
- Learning by Observation
- Final Presentations