BEGIN:VCALENDAR VERSION:2.0 PRODID:IEEE vTools.Events//EN CALSCALE:GREGORIAN BEGIN:VTIMEZONE TZID:America/Chicago BEGIN:DAYLIGHT DTSTART:20250309T030000 TZOFFSETFROM:-0600 TZOFFSETTO:-0500 RRULE:FREQ=YEARLY;BYDAY=2SU;BYMONTH=3 TZNAME:CDT END:DAYLIGHT BEGIN:STANDARD DTSTART:20251102T010000 TZOFFSETFROM:-0500 TZOFFSETTO:-0600 RRULE:FREQ=YEARLY;BYDAY=1SU;BYMONTH=11 TZNAME:CST END:STANDARD END:VTIMEZONE BEGIN:VEVENT DTSTAMP:20251017T142914Z UID:2C033AE4-7379-4E84-ADC1-4E7A1A41D371 DTSTART;TZID=America/Chicago:20251010T110000 DTEND;TZID=America/Chicago:20251010T120000 DESCRIPTION:[BME header](https://nam04.safelinks.protection.outlook.com/?ur l=https%3A%2F%2Ft.e2ma.net%2Fclick%2Fikehfk%2Fu9lp6f3%2Fa268qt&data=05%7C0 2%7Cmsol%40groute.uic.edu%7Cdc5abe6121dc4ce15fb008de06791fe9%7Ce202cd477a5 64baa99e3e3b71a7c77dd%7C0%7C0%7C638955314476440584%7CUnknown%7CTWFpbGZsb3d 8eyJFbXB0eU1hcGkiOnRydWUsIlYiOiIwLjAuMDAwMCIsIlAiOiJXaW4zMiIsIkFOIjoiTWFpb CIsIldUIjoyfQ%3D%3D%7C0%7C%7C%7C&sdata=43HVqdk%2BPQqgki4Oxe9vPAggO1%2B5MdA 5M%2FG3l0nyHMw%3D&reserved=0)\n\n[Image]\n\n-- ECE 595 Department Seminar\ n-- SPS Chicago Chapter Seminar\n\nFriday\, October 10th 11:00am - 12:30pm \n\nLecture Center C4\n\nDeep Learning for Smarter Algorithms in Detection \, Estimation\, and Navigation: From Underwater Environment Estimation to Robotic Path Planning\n\nToros Arikan\, PhD\n\nAssistant professor in Elec trical Engineering\n\nUniversity of Notre Dame\n\nAbstract: By deploying d eep learning methods as global optimizers that learn algorithmic solutions \, we can solve long-standing ill-posed problems in joint parameter estima tion. Neural networks can also produce human-interpretable results that de -mystify the algorithm design process\, allowing us to rigorously determin e hyperparameters such as training lengths. I will introduce two recent wo rks where this methodology is used to tackle open problems in remote sensi ng and navigation. In underwater environment estimation\, I will present a U-Net method for the acoustic mapping of reflective boundaries such as th e sea surface and seafloor\, with state-of-the-art performance and the new capability of jointly estimating the number of boundaries in the environm ent. In the field of robotic path planning\, I will present a recurrent co nvolutional neural network (RCNN) method for solving a sub-class of Obstac le Avoiding Rectilinear Steiner Minimum Tree (OARSMT) problems. By learnin g an algorithm via reinforcement learning\, whose intermediate stages are visible to the user\, we devise augmentations that yield strong accuracy a nd runtime performances\, which can lead to lower power consumption. Beyon d their immediate applications\, these new methods point to a general stra tegy of solving a broad class of joint parameter estimation problems via d eep learning.\n\nBio: Toros Arikan was born in Mount Kisco\, NY\, USA\, in 1993. He obtained his B.S. and M.S. degrees from the University of Illino is at Urbana-Champaign\, with a focus on digital communications and signal processing\; and his Ph.D. degree from the Massachusetts Institute of Tec hnology in 2023\, specializing in localization\, tracking\, and remote sen sing. In 2023-25 he was a postdoctoral researcher at Rice University\, whe re he researched deep learning methods for environment estimation\, and ro botic path planning algorithms for cluttered environments. He is currently a new assistant professor in EE at the University of Notre Dame\, where h e teaches Array Signal Processing and is actively recruiting graduate stud ents and postdoctoral candidates.\n\nFaculty Host: Dr. Ahmet Enis Cetin (a ecyy@uic.edu)\n\nRoom: C4\, Bldg: Lecture Center C4\, 802 S. Halsted St\, CHICAGO\, Illinois\, United States\, 60607 LOCATION:Room: C4\, Bldg: Lecture Center C4\, 802 S. Halsted St\, CHICAGO\, Illinois\, United States\, 60607 ORGANIZER:aecyy@uic.edu SEQUENCE:7 SUMMARY:IEEE SPS Chicago Chapter Seminar by Prof. Toros Arikan URL;VALUE=URI:https://events.vtools.ieee.org/m/506179 X-ALT-DESC:Description: <br /><table role="presentation" border="0" width=" 100%" cellspacing="0" cellpadding="0" align="center">\n<tbody>\n<tr>\n<td> \n<table class="m_2076634685384526817row-content m_2076634685384526817stac k" role="presentation" border="0" width="600" cellspacing="0" cellpadding= "0" align="center">\n<tbody>\n<tr>\n<td class="m_2076634685384526817column " width="100%">\n<table class="m_2076634685384526817image_block" role="pre sentation" border="0" width="100%" cellspacing="0" cellpadding="0">\n<tbod y>\n<tr>\n<td>\n<div align="center">\n<div class="m_2076634685384526817ful lWidth"><a href="https://nam04.safelinks.protection.outlook.com/?url=https %3A%2F%2Ft.e2ma.net%2Fclick%2Fikehfk%2Fu9lp6f3%2Fa268qt&amp\;data=05%7C02% 7Cmsol%40groute.uic.edu%7Cdc5abe6121dc4ce15fb008de06791fe9%7Ce202cd477a564 baa99e3e3b71a7c77dd%7C0%7C0%7C638955314476440584%7CUnknown%7CTWFpbGZsb3d8e yJFbXB0eU1hcGkiOnRydWUsIlYiOiIwLjAuMDAwMCIsIlAiOiJXaW4zMiIsIkFOIjoiTWFpbCI sIldUIjoyfQ%3D%3D%7C0%7C%7C%7C&amp\;sdata=43HVqdk%2BPQqgki4Oxe9vPAggO1%2B5 MdA5M%2FG3l0nyHMw%3D&amp\;reserved=0" target="_blank" rel="noopener" data- saferedirecturl="https://www.google.com/url?q=https://nam04.safelinks.prot ection.outlook.com/?url%3Dhttps%253A%252F%252Ft.e2ma.net%252Fclick%252Fike hfk%252Fu9lp6f3%252Fa268qt%26data%3D05%257C02%257Cmsol%2540groute.uic.edu% 257Cdc5abe6121dc4ce15fb008de06791fe9%257Ce202cd477a564baa99e3e3b71a7c77dd% 257C0%257C0%257C638955314476440584%257CUnknown%257CTWFpbGZsb3d8eyJFbXB0eU1 hcGkiOnRydWUsIlYiOiIwLjAuMDAwMCIsIlAiOiJXaW4zMiIsIkFOIjoiTWFpbCIsIldUIjoyf Q%253D%253D%257C0%257C%257C%257C%26sdata%3D43HVqdk%252BPQqgki4Oxe9vPAggO1% 252B5MdA5M%252FG3l0nyHMw%253D%26reserved%3D0&amp\;source=gmail&amp\;ust=17 60112260232000&amp\;usg=AOvVaw3vPKwv5fU5nXlwpg8QQK1C"><img class="CToWUd" title="BME header" src="https://ci3.googleusercontent.com/meips/ADKq_NZAdF a3xsMAQB8rxtoNpuCYeA2qvsMHkHM9uhSPZKXdPKkxSbixa9VxUrA9aiVNm892Fw8efcX6EDZv lnveqE45qSX1NQj63ZAClNIy93x5nsLT8B61UJEMnaIVOnfU7x2o7If2wTfo35UHYwQ=s0-d-e 1-ft#https://d31hzlhk6di2h5.cloudfront.net/20251008/4b/cb/d7/d4/99774560bb 4384dbfc2c4af5.jpg" alt="BME header" width="600" height="auto" data-bit="i it"></a></div>\n</div>\n</td>\n</tr>\n</tbody>\n</table>\n<table class="m_ 2076634685384526817image_block" role="presentation" border="0" width="100% " cellspacing="0" cellpadding="0">\n<tbody>\n<tr>\n<td>\n<div align="cente r">\n<div><img class="CToWUd a6T" tabindex="0" title="Image" src="https:// ci3.googleusercontent.com/meips/ADKq_NZn1l70CZyyHAprCnurXDSYm96mZufhz7wfUp NHBec-i3H6sGiene0t5EHrit-ZhfXQQ-mr2i3L_Xd3f0uQzxZ5Oyo2q4KM4OEv9561Rt_OPlP_ 2jZaIlaeSHCaOS8utPZZ8STtzvI9uwDO1qo=s0-d-e1-ft#https://d31hzlhk6di2h5.clou dfront.net/20251008/17/f1/3e/53/01752691a1d890427c8df00c.jpg" alt="Image" width="330" height="auto" data-bit="iit"></div>\n</div>\n</td>\n</tr>\n</t body>\n</table>\n<table role="presentation" border="0" width="100%" cellsp acing="0" cellpadding="20">\n<tbody>\n<tr>\n<td>\n<div>\n<div>\n<p><strong ><br><br>-- ECE 595 Department Seminar<br>-- SPS Chicago Chapter Seminar<b r><br></strong></p>\n</div>\n</div>\n</td>\n</tr>\n</tbody>\n</table>\n<ta ble role="presentation" border="0" width="100%" cellspacing="0" cellpaddin g="0">\n<tbody>\n<tr>\n<td>\n<div>\n<div>\n<p><strong>Friday\, October 10t h 11:00am - 12:30pm</strong></p>\n<p><strong>Lecture Center C4</strong></p >\n</div>\n</div>\n</td>\n</tr>\n</tbody>\n</table>\n<table role="presenta tion" border="0" width="100%" cellspacing="0" cellpadding="20">\n<tbody>\n <tr>\n<td>\n<div>\n<div>\n<p><strong>Deep Learning for Smarter Algorithms in Detection\, Estimation\, and Navigation: From Underwater Environment Es timation to Robotic Path Planning</strong></p>\n</div>\n</div>\n</td>\n</t r>\n</tbody>\n</table>\n</td>\n</tr>\n</tbody>\n</table>\n</td>\n</tr>\n</ tbody>\n</table>\n<table role="presentation" border="0" width="100%" cells pacing="0" cellpadding="0" align="center">\n<tbody>\n<tr>\n<td>\n<table cl ass="m_2076634685384526817row-content m_2076634685384526817stack" role="pr esentation" border="0" width="600" cellspacing="0" cellpadding="0" align=" center">\n<tbody>\n<tr>\n<td class="m_2076634685384526817column" width="10 0%">\n<table role="presentation" border="0" width="100%" cellspacing="0" c ellpadding="0">\n<tbody>\n<tr>\n<td>\n<div>\n<div>\n<p><strong>Toros Arika n\, PhD</strong></p>\n<p><strong>Assistant professor in Electrical Enginee ring</strong></p>\n<p><strong>University of Notre Dame</strong></p>\n</div >\n</div>\n</td>\n</tr>\n</tbody>\n</table>\n</td>\n</tr>\n</tbody>\n</tab le>\n</td>\n</tr>\n</tbody>\n</table>\n<table role="presentation" border=" 0" width="100%" cellspacing="0" cellpadding="0" align="center">\n<tbody>\n <tr>\n<td>\n<table class="m_2076634685384526817row-content m_2076634685384 526817stack" role="presentation" border="0" width="600" cellspacing="0" ce llpadding="0" align="center">\n<tbody>\n<tr>\n<td class="m_207663468538452 6817column" width="100%">\n<table role="presentation" border="0" width="10 0%" cellspacing="0" cellpadding="0">\n<tbody>\n<tr>\n<td>\n<div>\n<div>\n< p><strong>Abstract</strong>:&nbsp\;By deploying deep learning methods as g lobal optimizers that learn algorithmic solutions\, we can solve long-stan ding ill-posed problems in joint parameter estimation. Neural networks can also produce human-interpretable results that de-mystify the algorithm de sign process\, allowing us to rigorously determine hyperparameters such as training lengths. I will introduce two recent works where this methodolog y is used to tackle open problems in remote sensing and navigation. In und erwater environment estimation\, I will present a U-Net method for the aco ustic mapping of reflective boundaries such as the sea surface and seafloo r\, with state-of-the-art performance and the new capability of jointly es timating the number of boundaries in the environment. In the field of robo tic path planning\, I will present a recurrent convolutional neural networ k (RCNN) method for solving a sub-class of Obstacle Avoiding Rectilinear S teiner Minimum Tree (OARSMT) problems. By learning an algorithm via reinfo rcement learning\, whose intermediate stages are visible to the user\, we devise augmentations that yield strong accuracy and runtime performances\, which can lead to lower power consumption. Beyond their immediate applica tions\, these new methods point to a general strategy of solving a broad c lass of joint parameter estimation problems via deep learning.</p>\n</div> \n</div>\n</td>\n</tr>\n</tbody>\n</table>\n<table role="presentation" bor der="0" width="100%" cellspacing="0" cellpadding="0">\n<tbody>\n<tr>\n<td> \n<div>\n<div>\n<p><strong>Bio:&nbsp\;</strong>Toros Arikan was born in Mo unt Kisco\, NY\, USA\, in 1993. He obtained his B.S. and M.S. degrees from the University of Illinois at Urbana-Champaign\, with a focus on digital communications and signal processing\; and his Ph.D. degree from the Massa chusetts Institute of Technology in 2023\, specializing in localization\, tracking\, and remote sensing. In 2023-25 he was a postdoctoral researcher at Rice University\, where he researched deep learning methods for enviro nment estimation\, and robotic path planning algorithms for cluttered envi ronments. He is currently a new assistant professor in EE at the Universit y of Notre Dame\, where he teaches Array Signal Processing and is actively recruiting graduate students and postdoctoral candidates.&nbsp\;</p>\n<p> &nbsp\;</p>\n</div>\n</div>\n</td>\n</tr>\n</tbody>\n</table>\n<table role ="presentation" border="0" width="100%" cellspacing="0" cellpadding="0">\n <tbody>\n<tr>\n<td>\n<div>\n<div>\n<p>Faculty Host: Dr. Ahmet&nbsp\;<span class="il">Enis</span>&nbsp\;Cetin (<a href="mailto:aecyy@uic.edu" target= "_blank" rel="noopener">aecyy@uic.edu</a>)</p>\n</div>\n</div>\n</td>\n</t r>\n</tbody>\n</table>\n</td>\n</tr>\n</tbody>\n</table>\n</td>\n</tr>\n</ tbody>\n</table>\n<table role="presentation" border="0" width="100%" cells pacing="0" cellpadding="0" align="center">\n<tbody>\n<tr>\n<td>&nbsp\;</td >\n</tr>\n</tbody>\n</table> END:VEVENT END:VCALENDAR