<html><head></head><body><div class="yahoo-style-wrap" style="font-family:Helvetica Neue, Helvetica, Arial, sans-serif;font-size:13px;"><div dir="ltr" data-setdir="false"><div><strong><span style="font-size: 11.5pt; font-family: Arial, sans-serif; color: rgb(173, 39, 39);">29th Kanizsa Lecture</span></strong></div><div><strong><span style="font-size: 11.5pt; font-family: Arial, sans-serif; color: rgb(72, 79, 89);"><br></span></strong></div><div><strong><span style="font-size: 11.5pt; font-family: Arial, sans-serif; color: rgb(72, 79, 89);">Prof Michael Herzog</span></strong></div><div><span style="color: rgb(72, 79, 89); font-family: Arial, sans-serif; font-size: 11.5pt;">"THE IRREDUCIBILITY OF VISION: GESTALT, CROWDING AND THE FUNDAMENTALS OF VISION"</span></div><div><div><br></div><div>We are pleased to announce that the 29th Kanizsa Lecture will be delivered by Professor Michael Herzog. The event is open to the public. A limited number of guests will attend the lecture in the Sala Archivio Antico, Palazzo Bo, Padova, while everybody can attend the streamed live event online starting at 16:30 (UTC +2).</div><div><br></div><div>Youtube streaming: <a href="http://youtu.be/rpUoeLD85GE" rel="nofollow" target="_blank">http://youtu.be/rpUoeLD85GE</a> </div><div> </div><div>The event is also part of the celebrations organised as the University of Padova approaches its 800 anniversary, in the year 2022.</div><div> </div><div>Abstract</div><div> </div><div>What is fundamental to vision is controversially discussed since ancient times. The standard model of vision says that visual features are carefully analyzed in a feedforward hierarchy starting from low-level features, such as edges, to eventually compute objects. Low-level feature processing fully determines higher-level object appearance but not the other way around. Using crowding as an example, we show that this proposition does not hold true. Performance on a target deteriorates when flanking elements are presented, which is easily explained by low-level feature interactions in the standard model. However, adding further flankers can strongly improve the situation, depending on the overall configuration of the entire stimulus ensemble. Hence, contrary to the standard model, wholes determine parts as much as the other way around, very much in line with classic notions of Gestalt theory, which once was replaced by the standard model of vision. However, we did not succeed in fully explaining the effects of configuration by classic Gestalt explanations. One reason for this failure may be that, as we show, configural aspects are computed in a time dependent fashion. Based on these results and the ones of neuroimaging studies, we propose that the basic building blocks of vision are discrete processing windows, during which Gestalts are processed, which determine performance on its parts. The talk will start and end with some general philosophical considerations about vision.</div><div> </div><div>For more information:</div><div><a href="https://dpg.unipd.it/en/percup/kanizsa-lecture" rel="nofollow" target="_blank">https://dpg.unipd.it/en/percup/kanizsa-lecture</a> <br></div><div> </div><div>The organising committee:</div><div>Marco Bertamini, Paolo Bernardis, Luca Battaglini, Mario Bonato, Gianluca Campana, Carlo Fantoni, Walter Gerbino, Massimo Grassi</div><div dir="ltr" data-setdir="false"> </div></div></div></div></body></html>