Office of Naval Research: N000141010933

ONR MURI 2012-2013 Annual Meeting on

Knowledge Representation, Reasoning and Learning for Understanding Scenes and Events

[UCLA Royce Hall]

Session 1: Overview

8:15-8:25

Report on the Project theme and Progress in 2012-2013
Song-Chun Zhu

8:25-8:30

Comments from ONR
Behzad Kamghar-Parsi

Session 2: Object, Scene and Event Recognition

8:30-9:00

Large-scale learning of object shape and function
Deva Ramanan (UC Irvine)

9:00-9:30

3D scenes: SUN 3D database, and parsing IKEA furniture
Hamed Pirsiavash and Antonio Torralba (MIT)

9:30-10:00

Analyzing Three-Dimensional Scenes
Jitendra Malik (UC Berkeley)

10:15-10:30

Scene understanding – the mid-level soft underbelly
Michael Maire and Pietro Perona (Caltech)

10:30-10:45

Towards a causal analysis of behavior
Xavi Burgos and Pietro Perona (Caltech)

10:45-11:15

Scene Understanding by reasoning functionality and physics
Yibiao Zhao (UCLA)

11:15-12:00

Predicting Visual Memorability
Aude Oliva (MIT)

Session 3: Theories and Models

1:00-1:30

Unsupervised learning of generative image models: theory and applications
Yingnian Wu (UCLA)

1:30-2:00

Mini-Epitomes for Image Labeling and Image Modeling
Alan Yuille (UCLA)

2:00-2:30

Appearance Models
Stuart Geman (Brown)

2:30-3:00

Recent results on Transportability and Missing Data
Elias Barenboin, Karthika Mohan and Judea Pearl (UCLA)

Session 4: Knowledge Representation, Cognition, Reasoning and Queries

3:15-3:45

Simulation for physical scene understanding and social goal inference
Peter Battaglia, Tao Gao, and Josh Tenenbaum (MIT)

3:45-4:15

Crowd engineering and humans-in-the-loop learning
Jia Deng (Stanford)

4:15-4:45

Knowledge representation: ‘dark matters’, minds and dialogues
Song-Chun Zhu (UCLA)

Session 5: Panel Discussion

4:45-5:30

Challenges and Collaboration Plan