Select a calendar:
Filter January Events by Event Type:
Events for the 3rd week of January
-
PhD Thesis Defense - Meryem M'Hamdi
Tue, Jan 16, 2024 @ 10:00 AM - 12:00 PM
Thomas Lord Department of Computer Science
University Calendar
PhD Thesis Defense - Meryem MHamdi
Title: Towards More Human-Like Cross-lingual Transfer Learning
Committee Members: Jonathan May (Chair), Aiichiro Nakano, Khalil Iskarous.
Abstract: Cross-lingual transfer learning comprises a set of techniques used to adapt a model trained on (a) source language(s), enabling it to generalize to new target languages. With the emergence of Transformer-based contextualized encoders, there has been a surge in multilingual representations that adapt these encoders to various cross-lingual downstream applications. The surprising zero-shot capabilities of these encoders make them promising substitutes for other fully supervised techniques, bypassing the need for large-scale annotation. However, these representations are still far from solving the long-tail of NLP phenomenon, where models are biased more towards high-resource and typologically similar languages to the source language. This bias can be attributed to the over-reliance of current transfer learning pipelines on what we define as the 'Data-Intensive Identically-Distributed Minimally-Evaluated' paradigm. In this thesis, we analyze and propose techniques to advance the capabilities of multilingual language models beyond the traditional paradigm and more toward human-like cross-lingual transfer learning. We achieve that through 1) human-inspired input requirements by using few-shot meta-learning techniques, 2) human-inspired outcomes by defining a cross-lingual continual learning evaluation paradigm, and 3) human-inspired approaches through devising cognitive strategies to consolidate retention of knowledge learned across languages. Our contributions towards advancing the current transfer learning paradigm towards human-like learning are four-fold: 1) We explore cross-lingual fine-tuning on low-resource multilingual applications such as event trigger extraction and semantic search, shedding light on the strengths and limitations of existing cross-lingual transfer learning techniques. 2) We propose language-agnostic meta-learning approaches that can further bridge the gap between source and target typologically diverse languages. We show the merits of our approaches in reaching quicker and smoother generalization compared to naive fine-tuning, especially under low-resource scenarios. 3) We are the first to define a lifelong learning paradigm that analyzes language shifts. We show the merits and challenges of a multi-hop analysis where the system continually learns over several languages one at a time. 4) We are the first to adapt a cognitively inspired technique based on Leitner-queues to choose what to repeat in a cross-lingual continual learning setup and investigate its impact on reducing the forgetting of previously learned languages.Location: Henry Salvatori Computer Science Center (SAL) - 213
Audiences: Everyone Is Invited
Contact: CS Events
Event Link: https://usc.zoom.us/j/94477374759?pwd=ajZrd1o0QktVVXZsRk9UL1J6NGdtdz09#success
-
How to Develop an Effective Resume and Cover Letter
Tue, Jan 16, 2024 @ 12:00 PM - 01:00 PM
Viterbi School of Engineering Career Connections
Workshops & Infosessions
THIS EVENT WILL BE HOSTED HYBRID: IN-PERSON & ONLINE SIMULTANEOUSLY
Zoom link: https://usc.zoom.us/meeting/register/tJ0pce2pqzkjHN0WEJgvAEytXoJsY2BgIrDa
Does your resume highlight the skills that will land an interview? Review tips and tricks on how to best showcase yourself to employers by attending this professional development Q&A moderated by Viterbi Career Connection staff. Have your resume ready to take notes and make modifications in real time!
For more information about all workshops, please visit viterbicareers.usc.edu/workshops.Location: Ronald Tutor Hall of Engineering (RTH) - 211
Audiences: All Viterbi
Contact: RTH 218 Viterbi Career Connections
Event Link: https://usc.zoom.us/meeting/register/tJ0pce2pqzkjHN0WEJgvAEytXoJsY2BgIrDa
-
EiS Communications Hub Drop-In Hours
Wed, Jan 17, 2024 @ 10:00 AM - 01:00 PM
Viterbi School of Engineering Student Affairs
Workshops & Infosessions
Viterbi Ph.D. students are invited to stop by the EiS Communications Hub for one-on-one instruction for their academic and professional communications tasks. All instruction is provided by Viterbi faculty at the Engineering in Society Program.
Location: Ronald Tutor Hall of Engineering (RTH) - 222A
Audiences: Viterbi Ph.D. Students
Contact: Helen Choi
Event Link: https://sites.google.com/usc.edu/eishub/home?authuser=0
-
Computer Science General Faculty Meeting
Wed, Jan 17, 2024 @ 12:00 PM - 02:00 PM
Thomas Lord Department of Computer Science
Receptions & Special Events
Bi-Weekly regular faculty meeting for invited full-time Computer Science faculty only. Event details emailed directly to attendees.
Location: Hedco Neurosciences Building (HNB) - 107
Audiences: Invited Faculty Only
Contact: Assistant to CS Chair
-
Trojan Talk with myUniMate - Campus Recruitment
Wed, Jan 17, 2024 @ 03:00 PM - 03:45 PM
Viterbi School of Engineering Career Connections
Workshops & Infosessions
myUniMate is an early-stage pioneering student engagement platform designed to redefine the college journey. We believe that the student should be at the core of every experience. myUniMate is challenging the limits of generative AI and machine learning, for students to maximize their academic, social and career aspects of their university adventure!
We are seeking enthusiastic individuals who are motivated by ambitious goals to fulfill our mission. Ideal candidates thrive in a startup atmosphere, where challenging and unsolved problems are abundant. We value team members who are keen to learn, grow, and engage in collaborative efforts.
Open opportunities (permanent, contract, internships) exist in software engineering and development, conversation experience, product design and development, data science, marketing and more.
RSVP in ConnectSC events by clicking the “Attend” button
USC Virtual Career Session
Wednesday, 17 January 2024 · 3:00 – 4:00 pm
Video call link: https://meet.google.com/gsu-heea-sesOr dial: (US) 1 682-251-0844 PIN: 360 187 760Location: Virtual Event
Audiences: Everyone Is Invited
Contact: RTH 218 Viterbi Career Connections
Event Link: https://meet.google.com/gsu-heea-ses
-
AME Seminar - CANCELLED
Wed, Jan 17, 2024 @ 03:30 PM - 04:30 PM
Aerospace and Mechanical Engineering
Conferences, Lectures, & Seminars
Speaker: Rahul Panat, Carnegie Mellon University
Talk Title: Aerosol Jet Printed 3D Microarchitectures: From Process Development to Real-World Applications
Abstract: Please note: This seminar has been cancelled.
Host: AME Department
More Info: https://ame.usc.edu/seminars/
Webcast: https://usc.zoom.us/j/95892885119?pwd=QXZOZUhrcTJRYk5qZzZwVThrTytVZz09Location: James H. Zumberge Hall Of Science (ZHS) - 252
WebCast Link: https://usc.zoom.us/j/95892885119?pwd=QXZOZUhrcTJRYk5qZzZwVThrTytVZz09
Audiences: Everyone Is Invited
Contact: Tessa Yao
Event Link: https://ame.usc.edu/seminars/
-
A Guide to Viterbi Career Gateway: Your One-Stop Shop to Finding Jobs & Internships
Thu, Jan 18, 2024 @ 03:00 PM - 04:00 PM
Viterbi School of Engineering Career Connections
Workshops & Infosessions
THIS EVENT WILL BE HOSTED HYBRID: IN-PERSON & ONLINE SIMULTANEOUSLY
Zoom link: https://usc.zoom.us/meeting/register/tJIqf-ugrDkoGdRkOtjh88bYISlGKScEnOYz
Take part in a live tutorial to help you navigate Viterbi Career Gateway, a powerful job and internship search tool available ONLY to Viterbi students.
To access more information on all workshops, please visit viterbicareers.usc.edu/workshops.Location: Ronald Tutor Hall of Engineering (RTH) - 211
Audiences: Everyone Is Invited
Contact: RTH 218 Viterbi Career Connections
Event Link: https://usc.zoom.us/meeting/register/tJIqf-ugrDkoGdRkOtjh88bYISlGKScEnOYz
-
EiS Communications Hub Drop-In Hours
Fri, Jan 19, 2024 @ 10:00 AM - 01:00 PM
Viterbi School of Engineering Student Affairs
Workshops & Infosessions
Viterbi Ph.D. students are invited to stop by the EiS Communications Hub for one-on-one instruction for their academic and professional communications tasks. All instruction is provided by Viterbi faculty at the Engineering in Society Program.
Location: Ronald Tutor Hall of Engineering (RTH) - 222A
Audiences: Viterbi Ph.D. Students
Contact: Helen Choi
Event Link: https://sites.google.com/usc.edu/eishub/home?authuser=0
-
Medical Imaging Seminar: Louai Al-Dayeh, PhD - Practical Aspects of MRI Safety Test Methods of Active Implants
Fri, Jan 19, 2024 @ 10:00 AM - 11:30 PM
Ming Hsieh Department of Electrical and Computer Engineering
Workshops & Infosessions
Since the first successful MRI safety labeling of an implanted Deep Brain Stimulation (DBS) system approximately 20 years ago, active implantable medical device (AIMD) manufacturers have come a long way in designing their implants with MR safety in mind and in assessing what conditions of MR scanning (e.g., limits of RF and/or gradient) can allow MR imaging without compromising patient safety. MR Conditional implants undergo a wide range of well-developed test methods before receiving FDA approval under the specified conditions of use. These test methods include exposure in realistic MR imaging scanning environments, benchtop injection testing, and development of appropriate risk assessments though physical experiments and modeling. The seminar is an overview of the Practical Aspects of all MRI Safety Test Methods of active implants.
More Information: Louai Aldayeh_MHI-MISS.pdf
Location: Hughes Aircraft Electrical Engineering Center (EEB) - 132
Audiences: Everyone Is Invited
Contact: Mary Yung
-
ECE-EP Seminar - Jim Garrison - Friday, January 19th at 2pm in EEB 248
Fri, Jan 19, 2024 @ 02:00 PM - 03:30 PM
Ming Hsieh Department of Electrical and Computer Engineering
Conferences, Lectures, & Seminars
Speaker: Jim Garrison, Purdue University
Talk Title: Advancing the science and technology of Signals of Opportunity (SoOp) remote sensing
Series: ECE-EP Seminar
Abstract: Signals of Opportunity (SoOp) is an emerging field in microwave remote sensing in which existing anthropogenic signals (typically from communications or navigation satellites) are re-utilized in a non-cooperative manner as sources of illumination for bistatic radar. SoOp observations exhibit some properties common to either active radar or passive radiometry, but also have unique features distinct from these two classical approaches. Realizing the full potential of SoOp has required the development of new instruments, signal processing algorithms, geophysical model functions, and data assimilation methods. This presentation will review the fundamental theoretical and experimental research conducted by Prof. Garrison's group in these areas. SoOp signal models must integrate communication theory with the interaction between electromagnetic waves and natural media. Although many important geophysical variables measured by SoOp (e.g. ocean winds, soil moisture, and snow water equivalent) are the same as those observed by any other remote sensing technique, the basic electromagnetic quantities ("Level 1" data products) and their relationship to these geophysical variables are quite different. Direct assimilation of Level 1 data into Earth systems models, without explicitly inverting this relationship, could potentially reduce biases and improve their use in forecasting. Terrestrial and airborne campaigns are vital to this research both in the development of empirical model functions using in situ reference data and in the early-stage testing and demonstration of new instrument technologies. This talk will also highlight some potential pathways from fundamental research to application of SoOp remote sensing in Earth science missions, using three examples covering different stages of technical maturity. First, Global Navigation Satellite System Reflectometry (GNSS-R) is the most advanced SoOp technique. CYGNSS, launched in 2016, now has a large science community making use of various ocean, land and cryosphere variables extracted from its GNSS-R observations. Second, P-band (<400 MHz) communication signals exist in frequencies low enough to penetrate dense vegetation and soil, offering a capability for directly sensing Root-Zone Soil Moisture (RZSM). Prof. Garrison is the principal investigator on SNOOPI (SigNals Of Opportunity: P-band Investigation), a cubesat mission to be launched in Spring 2024 to demonstrate this technique. Finally, wide-band (~1GHz) communications signals in Ku-band (12-18 GHz) and higher can theoretically provide altimetry (sea surface height) at cm-level precision. A constellation of passive SoOp receivers could be launched for a fraction of the cost of a single active radar altimeter. Such a constellation could provide high temporal sampling of inland lakes and rivers for streamflow and discharge monitoring, and better coverage of coastal regions to observe rapidly evolving oceanographic features such as eddies.
Prof. Garrison will conclude the talk with some speculative concepts and ideas for future research directions.
Biography: James L Garrison received the B.S. degree from the Rensselaer Polytechnic Institute, Troy, NY, USA, the M.S. degree from Stanford University, Stanford, CA, USA, and the Ph.D. degree from the University of Colorado Boulder, Boulder, CO, USA, in 1988, 1990, and 1997, respectively. He is a Professor with the School of Aeronautics and Astronautics, Purdue University, West Lafayette, IN, USA, with a courtesy appointment at the School of Electrical and Computer Engineering. In 2022, he was elected a University Faculty Scholar. He made the first airborne measurements of ocean surface winds using reflected Global Navigation Satellite Systems (GNSS) signals in 1996 and continues to lead research in Earth remote sensing using signals of opportunity. He is the Principal Investigator for SNOOPI, a NASA mission to demonstrate remote sensing with P-band signals of opportunity. Prior to his academic position, he was with the National Aeronautics and Space Administration (NASA). Dr. Garrison is a fellow of both the Institute of Navigation (ION) and the IEEE. He served as Editor-in- Chief for the IEEE Geoscience and Remote Sensing Magazine from 2018 to 2022.
Host: ECE-Electrophysics
More Information: Jim Garrison Flyer.pdf
Location: Hughes Aircraft Electrical Engineering Center (EEB) - 248
Audiences: Everyone Is Invited
Contact: Marilyn Poplawski
-
PhD Dissertation Defense - Xuefeng Hu
Fri, Jan 19, 2024 @ 03:00 PM - 05:00 PM
Thomas Lord Department of Computer Science
University Calendar
PhD Dissertation Defense - Xuefeng Hu
Committee members: Ram Nevatia (chair), Aram Galstyan and Keith Jenkins
Title: Adapt Pre-trained Representation Towards Downstream Tasks
Abstract: In recent years, the field of computer vision and machine learning has witnessed a paradigm shift, characterized by a dramatic increase in the scale of model parameters and training data sizes. This evolution has led to significant enhancements in model accuracy and robustness, transcending the traditional, task-specific expert models. The field has now pivoted towards universal, large-scale pre-trained visual representations, which enables impressive zero-shot and few-shot solutions for a wide array of downstream tasks.
Despite these advancements, the application of pre-trained models to specific downstream tasks, each with their unique conditions and domain-specific challenges, often exposes inherent limitations. This dissertation aims to tackle these challenges. The research journey comprises a spectrum of approaches from fully-supervised to source-free and test-time adaptation, with diverse applications such as image classification, object detection, and forensic detection. This dissertation introduces novel architectures such as SPAN, which has pioneered the utilization of the self-attention mechanism in the field of computer vision, as well as innovative adaptation algorithms like ReCLIP and BaFTA, which enhance zero-shot classification performance with unsupervised vision-text alignment. This dissertation marks a transition from classic visual representations, like those used in ImageNet, to cutting-edge vision-language models like CLIP, and has overcome some of the most pressing challenges in the field.
The works of this dissertation play an important role in bridging the gap between generic visual representations and the specific, nuanced requirements of various real-world tasks. By doing so, it establishes new benchmarks in optimizing the performance of machine learning models in practical applications, reinforcing the role of advanced computational techniques in solving complex, real-world problems.
Zoom Link: https://usc.zoom.us/j/95935934090?pwd=RTFNcUorbndkaXA2UGtFWWkrbEtsUT09
Meeting ID: 959 3593 4090
Passcode: 442518
Location: Henry Salvatori Computer Science Center (SAL) - 213
Audiences: Everyone Is Invited
Contact: CS Events
Event Link: https://usc.zoom.us/j/95935934090?pwd=RTFNcUorbndkaXA2UGtFWWkrbEtsUT09