Linda Smith Colloquium - Shared screen with speaker view
Jesse Snedeker
Is it possible that we are overfitting our theories to aspects of typical experience that while true, are not critical to human development. Specifically, to the degree that development is similar in blind and sighted children, doesn't that place a strong constraint on the role of viewpoint in cognitive/language development. Similar arguments could be made about children who do not learn to walk or crawl etc
Daniel Kessler
I'd also be interested in how much time is spent looking at faces that are on screens (vs. real-world faces)
Tomer David Ullman
[Great questions. Unless otherwise specified I’m assuming these are for the Q&A at the end]
Brian Leahy
Their own hands, or others’ hands?
Brian Leahy
THat’s the answer
Jesse Snedeker
(can we ask that now bc I'm wondering that too)
Tomer David Ullman
Small Q for end: Is the interaction between hands and objects that of *grasping* or moving the objects? (This is related to a 2013 PNAS paper on ‘discovery of hands from motions’]
Shari (she/her) Liu
Should we think about these changes in headcam data as the consequence of learning, as the training data for learning, or both? [can skip if this is answered with subsequent data]
Jesse Snedeker
question for now: these are minutes seeing faces or bodies. But for face recognition variability of faces is probably critical input feature too. Do we know if the children in the two cultures are seeing equal variable face sets?
Brian Leahy
Tomer, my question was answered, thank you