April 2023 - Gradstudents - mailman.cs.vt.edu

Jie Bu's Ph.D. defense [Feb 9 at 4 pm]
by Anuj Karpatne 02 Dec '24

02 Dec '24

Hello everyone, You are cordially invited to the final defense of PhD student Jie Bu on Thursday, Feb 9 at 4 pm. Zoom Link: https://virginiatech.zoom.us/j/5142223333 Title: Achieving More with Less: Learning Generalizable Neural Networks With Less Labeled Data and Computational Overheads Committee: Dr. Anuj Karpatne (advisor) Dr. Adrian Sandu Dr. Ismini Lourentzou Dr. Naren Ramakrishnan Dr. Anish Arora (from Ohio State University) Abstract: Recent advancements in deep learning have demonstrated its incredible ability to learn generalizable patterns and relationships automatically from data in a number of mainstream applications. However, the generalization power of deep learning methods largely comes at the costs of working with very large datasets and using highly compute-intensive models. Many applications cannot afford these costs needed to ensure generalizability of deep learning models. For instance, obtaining labeled data can be costly in scientific applications, and using large models may not be feasible in resource-constrained environments involving portable devices. This thesis aims to improve efficiency in machine learning by exploring different ways to learn generalizable neural networks that require less labeled data and computational resources. We demonstrate that using physics supervision in scientific problems can reduce the need for labeled data, thereby improving data efficiency without compromising model generalizability. Additionally, we investigate the potential of transfer learning powered by transformers in scientific applications as a promising direction for further improving data efficiency through domain adaptation. On the computational efficiency side, we present two efforts for increasing parameter efficiency of neural networks through novel architectures and structured network pruning. Thanks, -Anuj ------------------- Anuj Karpatne Assistant Professor Computer Science, Virginia Tech https://people.cs.vt.edu/karpatne/

2 6

Room Available in Torgersen Hall for CS Graduate Students
by Mayo, Trey 29 Aug '23

29 Aug '23

A reminder for all that in spring 2022 we opened a room in Torgersen Hall available exclusively for CS Graduate Students. Torgersen Hall 3200 (TORG 3200) is located directly across from the CS Office of Graduate Programs (TORG 3210). Photos are attached. It provides space for students to meet as a group to work on projects or individuals who just need to a quiet place to study or hang out between classes. The room should not be used for personal/group meetings/parties or classes/tutoring sessions. It is not a reservable space. Access is available to any currently enrolled CS graduate student regardless of degree program or campus. Access to the room is controlled with your Hokie Passport (Student ID card). The room is available 24/7 as long as Torgersen Hall is open. If you have trouble accessing the room with your Hokie Passport, please visit the CS Office of Graduate Programs in TORG 3210 and we'll research the matter. (We believe we've resolved all previous access issues reported). Please note that this room is only for CS Graduate Students. If we find an abuse of the facilities, we will investigate and remove access. Please do your part as a responsible Hokie and help us keep the room clean and in good shape. We hope you enjoy. ============================================================== Trey Mayo, Ed.D. Director of Graduate Programs Virginia Tech | Department of Computer Science Torgersen Hall Suite 3210 620 Drillfield Drive (MC 0106) | Blacksburg, VA 24061 (540) 231-0780 | treymayo(a)vt.edu<mailto:treymayo@vt.edu> | http://cs.vt.edu<http://cs.vt.edu/>

1 2

Thank you for the success of The End of Semester event
by Mohannad Elhamod 29 Apr '23

29 Apr '23

Dear CS grad students, The CS grad council would like to thank all of you for making our End of Semester event a great success. We had one of the largest turn-outs in a very long time last Friday, and while it is unfortunate that we could not do the picnic at the park, it was still such a joy to see so many old and new faces and feel your positive and vibrant energy across the room. We also want to give a special shout out to @Wesley Woo <wesleymwoo(a)vt.edu> , @Sahar Heidari <heidari(a)vt.edu>, @Alexandra Thompson <alexthompson06(a)vt.edu>,and @Daniel Manesh <danielmanesh(a)vt.edu> for the amazing effort they put into organizing this event and taking care of all the details.The success of this event would not have been possible without them. Thank you all again for being a beautiful community, and we look forward to seeing you in future events! [image: Photo] Mohannad Elhamod PhD Candidate in Computer Science Vice President of CS Grad Council [image: phone-icon] elhamod(a)vt.edu [image: mobile-icon] wordpress.cs.vt.edu/elhamod [image: mobile-icon] Virginia Tech [image: twitter icon] <https://twitter.com/ElhamodMohannad> [image: linkedin icon] <https://www.linkedin.com/in/mohannad-elhamod-088b6118/>

1 0

Meghana Holla's MS Thesis Defense (Monday May 1st, 1:00-3:00pm)
by Ismini Lourentzou 29 Apr '23

29 Apr '23

Hello everyone, You are invited to join us for Meghana's MS thesis defense this Monday. *Title:* Commonsense for Zero-Shot Language Video Localization *Date:* Monday, May 1st *Time:* 1:00-3:00 pm *Location:* 3160A Torgersen Hall and Zoom https://virginiatech.zoom.us/j/87360587457 <https://www.google.com/url?q=https://virginiatech.zoom.us/j/87360587457&sa=…> *Committee:* Ismini Lourentzou (chair), Narendran Ramakrishnan, Lifu Huang *Abstract: *Zero-shot Natural Language-Video Localization (NLVL) is designed to train an NLVL model with only raw video data by dynamically generating video segment proposals and pseudo-query annotations. However, pseudo-queries generated by prior work are not fully grounded in the source video. Additionally, given the unstructured nature of the pseudo-queries, the video-query annotations lack common ground. To this end, we propose CORONET, a zero-shot NLVL framework that leverages commonsense information to bridge the gap between the video and generated pseudo-queries through a commonsense enhancement module. We employ Graph Convolution Networks (GCN) to encode commonsense information extracted from a knowledge graph, conditioned on the video, and cross-attention to enhance the encoded video and pseudo-query vectors prior to localization. Empirical evaluations on two datasets show that our model outperforms several zero-shot and weakly supervised baselines, highlighting the importance of commonsense reasoning in multimodal understanding. Best, Ismini -- *Ismini Lourentzou* <https://isminoula.github.io/> Assistant Professor Virginia Tech | Computer Science Email: ilourentzou(a)vt.edu

1 0

Re: [Gradstudents] Correction: [Bburg-fac] Uma Gunturi's MS Thesis Defense (Wednesday , May 3, 12:30pm)
by Eugenia Rho 29 Apr '23

29 Apr '23

Hi All, To clarify, the defense is on Wednesday and not Monday. The correct zoom link is: https://virginiatech.zoom.us/j/3141225773 Apologies for the confusion. On Sat, Apr 29, 2023, 10:06 AM Eugenia Rho <eugenia(a)vt.edu> wrote: > Hello all, > > > I would like to invite you to join us next Wednesday for Uma Gunturi’s > M.S. thesis defense. Below are the details. > > > Title: Linguistically Differentiating Acts and Recalls of Racial > Microaggressions on Social Media > > Date: Wednesday, May 3, 2023 > > Time: 12:30 - 2 pm > > Location: In-Person at Gilbert 1321 and over Zoom at > https://virginiatech.zoom.us/j/3141225773 > Committee: Eugenia H. Rho (Chair), Edward Fox, Ismini Lourentzou > > > Abstract: > > In this work, we examine the linguistic signature of online racial > microaggressions (acts) and how it differs from that of personal narratives > recalling experiences of such aggressions (recalls) by Black social media > users. We manually curate and annotate a corpus of acts and recalls from > in-the-wild social media discussions, and verify labels with Black workshop > participants. We leverage Natural Language Processing (NLP) and qualitative > analysis on this data to classify (RQ1), interpret (RQ2), and characterize > (RQ3) the language underlying acts and recalls of racial microaggressions > in the context of racism in the U.S. Our findings show that neural language > models (LMs) can classify acts and recalls with high accuracy (RQ1) with > contextual words revealing themes that associate Blacks with objects that > reify negative stereotypes (RQ2). Furthermore, overlapping linguistic > signatures between acts and recalls serve functionally different purposes > (RQ3), providing broader implications to the current challenges in content > moderation systems on social media. > > Thanks, > > Eugenia >

1 0

[Bburg-fac] Uma Gunturi's MS Thesis Defense (Monday , May 3, 12:30pm)
by Eugenia Rho 29 Apr '23

29 Apr '23

Hello all, I would like to invite you to join us next Wednesday for Uma Gunturi’s M.S. thesis defense. Below are the details. Title: Linguistically Differentiating Acts and Recalls of Racial Microaggressions on Social Media Date: Wednesday, May 3, 2023 Time: 12:30 - 2 pm Location: In-Person at Gilbert 1321 and over Zoom at https://virginiatech.zoom.us/j/85081753808 Committee: Eugenia H. Rho (Chair), Edward Fox, Ismini Lourentzou Abstract: In this work, we examine the linguistic signature of online racial microaggressions (acts) and how it differs from that of personal narratives recalling experiences of such aggressions (recalls) by Black social media users. We manually curate and annotate a corpus of acts and recalls from in-the-wild social media discussions, and verify labels with Black workshop participants. We leverage Natural Language Processing (NLP) and qualitative analysis on this data to classify (RQ1), interpret (RQ2), and characterize (RQ3) the language underlying acts and recalls of racial microaggressions in the context of racism in the U.S. Our findings show that neural language models (LMs) can classify acts and recalls with high accuracy (RQ1) with contextual words revealing themes that associate Blacks with objects that reify negative stereotypes (RQ2). Furthermore, overlapping linguistic signatures between acts and recalls serve functionally different purposes (RQ3), providing broader implications to the current challenges in content moderation systems on social media. Thanks, Eugenia

1 0

Updates to End of Semester Event (the event formerly known as the End of Semester BBQ)
by Wesley Woo 28 Apr '23

28 Apr '23

Hi folks, After looking forward to the weather forecast for the next few days (*spoiler alert*: lots of rain), we've decided to turn the End of Semester BBQ into an *indoor* *End of Semester Pizza Party*! We will be eating pizza from *5 to 8 PM in Torg 1100* and there will also be board games and other fun activities that are like board games, but different. In case your availability has changed, I've updated the RSVP to include a "Not attending" option -- please update your response so we have an accurate headcount, and order the right amount of pizza! Updated RSVP link: https://forms.gle/bz4ViSo8YJAwSRFv9 <https://forms.gle/bz4ViSo8YJAwSRFv9> Looking forward to seeing everyone on Friday! Wesley

1 1

CS|SOURCE - EMPLOYER JOB POSTINGS AND EVENTS
by Julia Costello 28 Apr '23

28 Apr '23

CS|SOURCE - EMPLOYER JOB POSTINGS AND EVENTS Current opportunities with employers are listed below. View other positions in <https://career.vt.edu/job-search/Handshake.html> Handshake. OPPORTUNITY 1 <https://www.costargroup.com/> CoStar Group technology leaders will meet & greet students in the CS student lounge, Torgersen 1070 on Monday, May 1, 11am-1:30pm, Torgersen 1070. Flyer attached. OPPORTUNITY 2 Graduating Seniors - Career and Professional Development encourages graduating seniors to complete their First Destination Survey. This information is helpful to many people for many and varied reasons. If you do not have plans at the moment, that's ok (and not uncommon). You can let CPD know that in the survey and they'll reach out again in 6 months. We appreciate your assistance in sharing your information. Flyer attached. OPPORTUNITY 3 Atlanta Hawks seek a Data Analyst intern this summer. The position is remote. For more info and to apply, click: <https://wd1.myworkdaysite.com/recruiting/hawks/External/job/Atlanta-GA/Inte rn--Data-Analyst_00069> https://wd1.myworkdaysite.com/recruiting/hawks/External/job/Atlanta-GA/Inter n--Data-Analyst_00069. 2023 CORPORATE STAKEHOLDERS Julia Costello Strategic Partnership Coordinator Virginia Tech | College of Engineering | Department of Computer Science 2202 Kraft Drive | Blacksburg, VA 24060 540.250.6121 | <mailto:juliapc7@cs.vt.edu> juliapc7(a)cs.vt.edu CS|Source Website: <https://gcc02.safelinks.protection.outlook.com/?url=https%3A%2F%2Fcs.vt.edu %2Fcs-source.html&data=04%7C01%7Cgcaufman%40peraton.com%7C274bbb94bdea4a9c7b 0408d8e8aaee69%7C1c6258462b0a448383dde024820875b3%7C0%7C0%7C6375151681480464 16%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1ha WwiLCJXVCI6Mn0%3D%7C1000&sdata=VBHHIabAuZqQzfNpOVOGp0RpKf9LGzMs7qY3UNy3pm0%3 D&reserved=0> https://cs.vt.edu/cs-source.html Stay up to date on CS|Source job and event opportunities <https://cs.vt.edu/cs-source.html> CS|Source Website | <https://www.cs-source.cs.vt.edu/> CS|Source Portal | <https://twitter.com/vt_CS_Source> CS|Source on Twitter

1 0

Research Excellence Award
by Cal Ribbens 28 Apr '23

28 Apr '23

CS Folks, Please join me in congratulating MS student Meghana Holla who received the Paul E. Torgersen Research Excellence Award from the College of Engineering last week. Meghana is advised by Ismini Lourentzou. Way to go, Meghana! Dr. Ribbens -- Head, Department of Computer Science Virginia Tech https://cs.vt.edu/people/faculty/cal-ribbens.html https://twitter.com/CalRibbens_VTCS

2 2

Zoom Link for Matthew Jackson's M.Sc. Defense
by Dimitrios Nikolopoulos 28 Apr '23

28 Apr '23

Dear all, We provide a Zoom link for Matthew's defense, if you wish to attend remotely. Details below. The in-person meeting is in Room 3211, Gilbert Place at 11.30 am. Matthew Jackson is inviting you to a scheduled Zoom meeting. Topic: Matthew Jackson's Thesis Defense Time: Apr 28, 2023 11:30 AM Eastern Time (US and Canada) Join Zoom Meeting https://virginiatech.zoom.us/j/83013393393 Meeting ID: 830 1339 3393 One tap mobile +13126266799,,83013393393# US (Chicago) +16469313860,,83013393393# US Dial by your location +1 312 626 6799 US (Chicago) +1 646 931 3860 US +1 929 436 2866 US (New York) +1 301 715 8592 US (Washington DC) +1 305 224 1968 US +1 309 205 3325 US +1 386 347 5053 US +1 507 473 4847 US +1 564 217 2000 US +1 669 444 9171 US +1 669 900 6833 US (San Jose) +1 689 278 1000 US +1 719 359 4580 US +1 253 205 0468 US +1 253 215 8782 US (Tacoma) +1 346 248 7799 US (Houston) +1 360 209 5623 US Meeting ID: 830 1339 3393 Find your local number: https://virginiatech.zoom.us/u/kjPDBGYwe Join by SIP 83013393393(a)zoomcrc.com Join by H.323 162.255.37.11 (US West) 162.255.36.11 (US East) 221.122.88.195 (China) 115.114.131.7 (India Mumbai) 115.114.115.7 (India Hyderabad) 213.19.144.110 (Amsterdam Netherlands) 213.244.140.110 (Germany) 103.122.166.55 (Australia Sydney) 103.122.167.55 (Australia Melbourne) 209.9.211.110 (Hong Kong SAR) 149.137.40.110 (Singapore) 64.211.144.160 (Brazil) 149.137.68.253 (Mexico) 69.174.57.160 (Canada Toronto) 65.39.152.160 (Canada Vancouver) 207.226.132.110 (Japan Tokyo) 149.137.24.110 (Japan Osaka) Meeting ID: 830 1339 3393 Join by Skype for Business https://virginiatech.zoom.us/skype/83013393393 I invite you to attend Matthew Jackson's Master's thesis examination on Friday, April 28 2023, at 11:30 am. The defense will take place in Room 3211 at Gilbert Place. Details are listed below. Title: Computational Offloading for Real-Time Computer Vision in Unreliable Multi-Tenant Edge Systems Abstract: The demand and interest in serving Computer Vision applications at the Edge, where Edge Devices generate vast quantities of data, clashes with the reality that many Devices are largely unable to process their data in real-time. While computational offloading, not to the Cloud but to nearby Edge Nodes, offers convenient acceleration for these applications, such systems are not without their constraints. As Edge networks may be unreliable or wireless, offloading quality is sensitive to communication bottlenecks. Unlike seemingly unlimited Cloud resources, an Edge Node, serving multiple clients, may incur delays due to resource contention. This project describes relevant Computer Vision workloads and how an effective offloading framework must adapt to these constraints that are a key impact to Quality of Service yet has not been effectively nor properly addressed by previous literature. We design an offloading controller, based on closed-loop control theory, that gives Devices the ability to maximize their throughput by appropriately offloading under variable conditions. This approach ensures a Device can utilize the maximum available offloading bandwidth. Finally, we constructed a realistic testbed and conducted measurements to demonstrate the superiority of our offloading controller over previous techniques. Committee: Dimitrios S. Nikolopoulos (Chair) Bo Ji (Co-Chair) Liting Hu (Member) We are looking forward to seeing many of you there. -- Dimitrios S. Nikolopoulos John W. Hancock Professor of Engineering Professor, Department of Computer Science Professor (by courtesy), Department of Electrical and Computer Engineering Associate Director, Stack@CS Center for Computer Systems Research dsn(a)vt.edu | x14260

1 0