BEGIN:VCALENDAR VERSION:2.0 PRODID:IEEE vTools.Events//EN CALSCALE:GREGORIAN BEGIN:VTIMEZONE TZID:America/New_York BEGIN:DAYLIGHT DTSTART:20210314T030000 TZOFFSETFROM:-0500 TZOFFSETTO:-0400 RRULE:FREQ=YEARLY;BYDAY=2SU;BYMONTH=3 TZNAME:EDT END:DAYLIGHT BEGIN:STANDARD DTSTART:20211107T010000 TZOFFSETFROM:-0400 TZOFFSETTO:-0500 RRULE:FREQ=YEARLY;BYDAY=1SU;BYMONTH=11 TZNAME:EST END:STANDARD END:VTIMEZONE BEGIN:VEVENT DTSTAMP:20210519T171703Z UID:6B913B1B-8F0A-487A-B24E-C6EA8EAB8A13 DTSTART;TZID=America/New_York:20210519T120000 DTEND;TZID=America/New_York:20210519T130000 DESCRIPTION:In this talk\, I first give an overview of the state-of-the-art development in deep learning and AI. Then\, I present two lines of our re cent work: (i) Grammar Guided Representation Learning: Grammar models are natural\, interpretable and fundamental schema in both language and image representation learning. Can grammars help us design better deep machine l earning models that are interpretable and parsimonious? I will discuss two ways of harnessing the best of the two worlds\, Grammars and Deep Neural Networks (DNNs): using a simple 1-D grammar to rethink and unify the desig n of neural architectures that achieve state-of-the-art results in compute r vision\, and using a simple 2-D grammar to rationalize state-of-the-art DNN-based object detection systems. (ii) Deep Consensus Learning: Both gen erative learning and discriminative learning have recently witnessed remar kable progress using DNNs. For structured input synthesis and structured o utput prediction problems (e.g.\, layout-to-image synthesis and image sema ntic segmentation respectively)\, they often are studied separately. As an example\, I will discuss deep consensus learning (DCL) for joint layout-t o-image synthesis and weakly-supervised image semantic segmentation. The p roposed DCL sheds light on exploiting DNNs for distilling structured knowl edge in many other domains. To conclude\, I will also discuss some thought s on building trustworthy and responsible AI under broader settings.\n\nSp eaker(s): Tianfu (Matt) Wu\, \n\nAgenda: \nMechatronics and Automation for Revitalizing Critical infrastructure Health (MARCH) - Moving from Researc h to Application\n\nThe IEEE Industrial Electronics Society Chapter of the Eastern North Carolina Section is hosting a multi-disciplinary webinar se ries titled\, “Mechatronics and Automation for Revitalizing Critical inf rastructure Health (MARCH) - Moving from Research to Application”. The s peakers will be focusing on applied research and development in automation of sensors\, actuators and intelligent decision making to improve\, autom ate or smarten critical infrastructure systems such as power\, communicati on\, water\, transportation\, etc. for fault detection\, diagnosis and mit igation\, and interconnected collaborative operation. The goal of the webi nar series is to build and expand connections between faculty and to foste r new faculty-industry relationships to widen the applicability of their r esearch to a larger industry/academic audience. This webinar also aims to broaden the research horizon and facilitate the integration of scientific aspects that were previously not considered as prominent factors for the o peration of certain critical infrastructure. Attendees will have the oppor tunity to interact with the speakers and other attendees to build partners hips to participate in funding opportunities and collaborative interdiscip linary research.\n\nRaleigh\, North Carolina\, United States\, Virtual: ht tps://events.vtools.ieee.org/m/271712 LOCATION:Raleigh\, North Carolina\, United States\, Virtual: https://events .vtools.ieee.org/m/271712 ORGANIZER:bbalago@ncsu.edu SEQUENCE:2 SUMMARY:Webinar Series - MARCH - Grammar Guided Representation Learning and Deep Consensus Learning URL;VALUE=URI:https://events.vtools.ieee.org/m/271712 X-ALT-DESC:Description: <br /><p>In this talk\, I first give an overview of the state-of-the-art development in deep learning and AI. Then\, I presen t two lines of our recent work: (i) Grammar Guided Representation Learning :&nbsp\; Grammar models are natural\, interpretable and fundamental schema in both language and image representation learning. Can grammars help us design better deep machine learning models that are interpretable and pars imonious? I will discuss two ways of harnessing the best of the two worlds \, Grammars and Deep Neural Networks (DNNs): using a simple 1-D grammar to rethink and unify the design of neural architectures that achieve state-o f-the-art results in computer vision\, and using a simple 2-D grammar to r ationalize state-of-the-art DNN-based object detection systems.&nbsp\; (ii ) Deep Consensus Learning: Both generative learning and discriminative lea rning have recently witnessed remarkable progress using DNNs. For structur ed input synthesis and structured output prediction problems (e.g.\, layou t-to-image synthesis and image semantic segmentation respectively)\, they often are studied separately. As an example\, I will discuss deep consensu s learning (DCL) for joint layout-to-image synthesis and weakly-supervised image semantic segmentation. The proposed DCL sheds light on exploiting D NNs for distilling structured knowledge in many other domains. To conclude \, I will also discuss some thoughts on building trustworthy and responsib le AI under broader settings.</p><br /><br />Agenda: <br /><p>Mechatronics and Automation for Revitalizing Critical infrastructure Health (MARCH) &n bsp\;- Moving from Research to Application</p>\n<p>The IEEE Industrial Ele ctronics Society Chapter of the Eastern North Carolina Section is hosting a multi-disciplinary webinar series titled\, &ldquo\;Mechatronics and Auto mation for Revitalizing Critical infrastructure Health (MARCH)&nbsp\; - Mo ving from Research to Application&rdquo\;. The speakers will be focusing o n applied research and development in automation of sensors\, actuators an d intelligent decision making to improve\, automate or smarten critical in frastructure systems such as power\, communication\, water\, transportatio n\, etc. for fault detection\, diagnosis and mitigation\, and interconnect ed collaborative operation. The goal of the webinar series is to build and expand connections between faculty and to foster new faculty-industry rel ationships to widen the applicability of their research to a larger indust ry/academic audience. This webinar also aims to broaden the research horiz on and facilitate the integration of scientific aspects that were previous ly not considered as prominent factors for the operation of certain critic al infrastructure.&nbsp\;Attendees will have the opportunity to interact w ith the speakers and other attendees to build partnerships to participate in funding opportunities and collaborative interdisciplinary research.</p> END:VEVENT END:VCALENDAR