Proud to highlight the hard work of this amazing team of researchers in the Harvard Vision Sciences Lab! #VSS2020. 1/n
@fenildoshi009 explores how to link neurophysiology to perceptual psychology using deep neural networks. Poster:rb.gy/vfo0sw Video Walkthrough:rb.gy/gpmeix 2/n 
@catmag29 finds transitions in the nature of the representational space along the ventral stream, leveraging different behavioral tasks. Talk: rb.gy/gw7fhw 3/n
@emiliejosephs introduces the first large-scale database of images depicting reachable environments. Poster: rb.gy/lpfzot Video Walkthrough: rb.gy/guve71 4/n
@jacob_s_prince presents a case for an integrated view of the ventral visual stream, based on empirical data relating category-selective responses in human brains and deepnets. Video: rb.gy/uh2vmn 5/n
@jeongho__park examines how the object-scene continuum is represented in the brain. Poster:rb.gy/jqx4nh Video Walkthrough:rb.gy/frdt9a 6/n
Colin Conwell digs into claims that rodent visual cortex is just a randomly initialized network. Poster: rb.gy/lpzuq0 Walkthrough: 7/n
@wang_ruosi probes the time course of mid-level feature processing with texform EEG decoding! Poster: rb.gy/urjvox Video Walkthrough: rb.gy/hu9foz 8/n
@DanJanini finds human-like representations of approximate number in neural networks trained to do object categorization. Poster: rb.gy/czpdhy Walkthrough: rb.gy/yecteh 9/n
@AylinKallmayer explores representations that support object, scene, and face recognition using deepnet trajectory analysis. Poster: rb.gy/jhucxh, video: rb.gy/ktxud7 10/n
@johnmark_taylor finds that CNNs trained for object recognition encode color/shape combinations in an increasingly conjoined, interactive format throughout processing Poster: rb.gy/ydsu5w Video: rb.gy/nwe0gr 11/n
@ArtDeza examines what foveation can do for scene representation using deep neural networks. Poster: rb.gy/wrvxjv Video Walkthrough: rb.gy/yvchyf Preprint: rb.gy/k3swhq 12/n
@JulianDeFreitas asks: can you have multiple self representations at once? Poster: rb.gy/3wlgl3. Video walkthrough: rb.gy/fnoyhn. 13/n
@LYTarhan and @JulianDeFreitas find that semantic features derived from natural language processing models predict intuitive judgments about how similar everyday actions are to one another. Poster: rb.gy/8jqagb Video walkthrough: 14/n
@_YiChiaChen_ asks why we like to see some objects big on the screen and others small--and shows these preferences do not depend on recognizing the object! Poster: rb.gy/x2qwpx Walkthrough: 15/n
@Hrag_P uses tDCS to identify separate neural substrates constraining VWM storage vs manipulation; then enhances each ability by up to 26% Talk: 16/n
@gcaedwards finds prolonged attention to one visual field boosts attention processing in the other visual field. Video walkthrough: 17/n
William Schmitt and @Hrag_P use tRNS to enhance the encoding of information and create efficient representations in VWM. Poster: rb.gy/iaedek Walkthrough: rb.gy/7tvj4t 18/n
@talia_konkle and @grez72 trained deepnets without any category labels, and found they match human brain responses to objects as well or better than supervised models. Video: rb.gy/ouimxn Preprint: rb.gy/m7im8b, Poster: rb.gy/7p3d5p 19/n
Don't hesitate to reach out to any of us with questions or thoughts, or just to say hi! Hope to chat with many of you at V-VVS. 20/20. (<== wha?!! surprisingly apt tweet count :)
