Lecture Notes - Michelle Greene
- Human eye: 126 million photoreceptors
- The optics of the eye are very bad
- When we walk, most of the time we stare right at the ground
- We think we see the full picture of what is around us
- Scene structure influences scene motion
- Depth of the scene gives us meaning
- Linear perspective changes the things we can do or say
- Global property space: concealment, temperature, expansiveness, openness, navigability,
transience, mean depth
- Computational, rapid perception, selective adaptation, influences categorization,
neural responses
- Predictions: more similarity, more false alarms
- JJ Gibson: “The meaning or value of a thing is what it affords”
-US Bureau of Labor Statistics → how do people spend their time?
- 227 actions
- 17 major categories
- Affordances best match human scene representations
- below the ceiling⅔
- Affordance must explain most unique variance
-Attributes to the world → sunny, busy, driving, walking (features in an environment)
- Research plan: extract variety of scene and decorrelate them
- EEF on the brain; millisecond temporal resolution, centimeter spatial resolution, direct
measure of neural populations, best at superficial sources
- EPR encoding for brain response
- Results: behavior predicting ERPs
-1st peak → posterior electrodes, 2nd peak → anterior electrodes
- Shared variance between features and behavior
- Is vision is primarily bottom-up, then the probability should not matter
- Learning to love the questions: be patient and persistent, read, get a great team
