Compositional Models - Complexity of Representation and Interference

Compositional Models - Complexity of Representation and Interference

Explore the intricacies of compositional models in this 51-minute lecture by Alan Yuille from the University of California, Los Angeles. Delve into the complexity of representation and interference, covering topics such as data-driven models, feedforward and feedback mechanisms, generative models, and hierarchies. Examine the history of generative models and their application to cluttered data scenarios. Analyze cross-examples, parameters, and local ambiguity in various model types. Gain insights into the importance of hierarchical structures in compositional modeling through in-depth discussions and examples.