BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Memento EPFL//
BEGIN:VEVENT
SUMMARY:Perception-Action Cycle: Implications of Active Learning for Devel
 oping Visual Intelligence
DTSTART:20210824T153000
DTEND:20210824T173000
DTSTAMP:20260407T021040Z
UID:5077b8895f9565e071c02e6f21e7c7859a77cbf700e50cd8a50f9904
CATEGORIES:Conferences - Seminars
DESCRIPTION:Onur Beker\nEDIC candidacy exam\nexam president: Prof. Volkan 
 Cevher\nthesis advisor: Prof. Amir Zamir\nco-examiner: Prof. Mackenzie Mat
 his\n\nAbstract\nEven simple animals like insects\, fish\, or frogs have a
  remarkable visual capacity to robustly navigate\, interact with\, and sur
 vive in the real world. And although there is a significant body of experi
 mental work aimed at discovering the general principles underlying these c
 apacities\, it is yet unclear how the brain decodes visual stimuli to high
 -level representations (e.g. objects and events) that can guide decisions 
 and actions. Similarly\, the standard task-centric approaches in computer 
 vision (e.g. detecting objects\, computing optic flow or building depth ma
 ps)\, while definitely useful\, are not yet sufficient to enable autonomy 
 and adaptive behavior in complex real-world environments\, at a level of f
 lexibility comparable to that of natural intelligence. An alternative dire
 ction is to instead adopt an agent-centric approach to the problem\, and c
 onduct an exploratory computational study of active vision by iteratively 
 optimizing autonomous systems on down-stream tasks that require a non-triv
 ial visual capacity for successful completion\, and analyzing the implicat
 ions of this optimization procedure on the perception models of the said s
 ystems. This way\, neither vision nor active behavior is treated as a stan
 d-alone function\, and the relationship between them is instead formulated
  as that of co-dependent participants with reciprocal contributions in a p
 erception-action cycle\, rather then that of a benefactor serving a benefi
 ciary.\n\nBackground papers\n\n	A Sensorimotor Account of Vision and Visua
 l Consciousness \n	Learning Generalizable Visual Representations via Inte
 ractive Gameplay \n	Proximal Policy Optimization Algorithms\n
LOCATION:
STATUS:CONFIRMED
END:VEVENT
END:VCALENDAR