Miguel Lázaro-Gredilla

This paper introduces a model called CMRF that learns attention-controllable border-ownership representations for perceptual grouping and object inference in novel scenes.

The paper develops a cortical circuit model based on a generative vision model to explain visual phenomena and functional roles of neural connections.

This paper introduces a cognitive graph structure called CSCG to model how the hippocampus forms abstract cognitive maps, explaining various cognitive phenomena and place cell behaviors.

This paper introduces a cloned HMM to model hippocampal cognitive maps as higher-order graphs, showing it captures key properties of hippocampal cells in rodents and humans.

This paper introduces a biologically plausible learning algorithm for Cloned HMMs that enables rapid memorization and gradual generalization of higher-order sequential structures.

This paper uses the Recursive Cortical Network to model and explain visual cortex phenomena like subjective contours and border-ownership competition.

The paper derives cortical circuit models based on a generative vision model, assigning computational roles to different connections and pathways.