Overview
Explore the intricacies of recurrent computations in the cortex through this comprehensive lecture by Gabriel Kreiman from Harvard University and Children's Hospital Boston. Delve into topics such as image captioning, visual routines, and macaque connectivity while examining deep completion network models. Gain insights into the human visual system, neuronal preferences, and the concept of backward masking. Investigate building models, desiderata, and various example questions to deepen your understanding of cortical computations and their implications for neuroscience and artificial intelligence.
Syllabus
Intro
Image Captioning
Example Questions
Building Models
Desiderata
Visual Routines
macaque connectivity
deep completion network models
how do we know
human visual system
what neurons really want
study
backward masking
Taught by
MITCBMM