Projects

Conceptulization with stochastic sets: regimes, transitions and composition

1. Atomic stochastic sets: the regimes of Gibbs models (high entropy) and Sparsity models (low Entropy)
2. The spectrum of entropy regimes: information scaling and model transition
3. Conceptualization by composition and stochastic grammar (middle entropy)

Representation: probabilistic And-or Graphs

1. S-AoG: representing the spatial hierarchy---scenes, objects, parts and primitives.
   such as, indoor and outdoor scenes, Human poses, Vehicles, and attributes.
2. T-AoG: representing the temporal hierarchy---event, action and motion primitive.
   such as human vehicle interactions, group activities in surveillance video
3. Summary: unification on visual knowledge representation and reasoning

Inference: goal guided, cost sensitive computing

1. The alpha, beta, gamma and omega processes in the AoG inference
2. Optimal decision policy
3. Tracking and online learning
4. Efficiency of sensors and features for nodes in the AoG
5. Scheduling the computational process: exploiting sensor efficacy and utility
6. SEE Inference in video: PTZ tracking guided by utility, sensor efficacy and AoG

Theory of Learning: information projection, learnability and generalizability

1. A unified learning theory based on information projection
2. PAC-learning of the inductively learned sets for nodes in the AoG
3. Generalizability: capacity and transportability of the S/T/C-AoGs
4. Curriculum learning and energy landscapes

Text parsing and probabilistic query answering

1. Lexical and syntactic processing
2. Semantic tagging and information integration for Joint Parse Graph
3. Text query parsing and updating utility functions from queries
4. Evaluation tool for text query

Airborne imagery registration and understanding

1. Geo-registration of WAMI and MAMI images
2. Target tracking in Airborn images
3. Integrating parse graphs from Airborne and ground sensors