Our lab is ready for Virtual-@VSSMtg! Check out these talks and posters by @ChenxiaooooGuan, @AlonHafri, @LittlestPat, @neuro_zz, @NartkerMakaela, Michael Lepori, Zekun Sun, Qian Yu, and special faculty guest Justin Halberda! Here's a bit more about what's on the menu. #VSS2020 
Zekun has discovered a "Goldilocks" relationship between complexity & beauty. She generated shapes whose informational complexity could be measured objectively, then had subjects decide which makes a better "painting". Surprisingly, *moderately* complex ones were best! #VSS2020
Justin, Qian, and Zekun continue the Goldilocks theme by exploring how complexity drives visual search & working memory. Lots of work suggests a connection between the two, but we suggest that the relationship may be quadratic! #VSS2020
We can appreciate not only how the world looks now, but also how it might look some other time—as when we see how disconnected puzzle pieces could combine into a complete whole. Does this affect other kinds of perceptual processing? @ChenxiaooooGuan says YES: numerosity! #VSS2020
@AlonHafri connects a scene's "physics" to attentive tracking of objects within it. He revisits perhaps the most foundational finding in visual cognition & shows that it is sensitive to physical relations in previously unknown ways. *Relations* matter for vision science! #VSS2020
We can represent objects that are completely covered by cloths or drapes (as in a recent PNAS paper by @flipphillips and Roland Fleming). Does this happen rapidly and automatically? @LittlestPat says yes! Draped objects "prime" their uncovered counterparts. #VSS2020
Intelligent machines rival "human-level" performance at classifying natural images. But sometimes they make big and surprising mistakes. Can humans anticipate when this will happen? And even for natural, unmodified images?
@neuro_zz and @NartkerMakaela say yes! #VSS2020
@neuro_zz and @NartkerMakaela say yes! #VSS2020
We often show each other visual examples—eg when we point out a ball to a dog or child, or teach someone how to spot poison ivy. Can people select good training examples? @NartkerMakaela & Mikey Lepori show that they can, using machine learning to study human "teaching"! #VSS2020
