Select a calendar:
Filter January Events by Event Type:
SUNMONTUEWEDTHUFRISAT
Conferences, Lectures, & Seminars
Events for January
-
CS Colloquium: Filip Ilievski (USC ISI) - Open-world AI Agents with Common Sense
Tue, Jan 11, 2022 @ 11:00 AM - 12:00 PM
Computer Science
Conferences, Lectures, & Seminars
Speaker: Filip Ilievski, USC
Talk Title: Open-world AI Agents with Common Sense
Abstract: In this talk, I will present my research on building natural-language AI agents that use common sense to act robustly in everyday situations. State-of-the-art technology is inadequate for this purpose: background knowledge and rules provide explainability but they cannot generalize to unseen situations, whereas neural models with natural generalizability are prone to making silly mistakes and are unable to explain their decisions. I will describe our agents which combine the best of both worlds to perform well on open-world tasks without the need for task-specific training data, by 1) organizing commonsense knowledge into a single harmonized resource; 2) enriching commonsense knowledge with axioms and preconditions; and 3) using the enriched knowledge resource to adapt neural models for out-of-domain scenarios that require common sense. The effectiveness of the combined approach is demonstrated through high performance on a range of out-of-domain tasks, including question answering and story generation. I will also reflect on my ongoing research on adapting AI agents to specific domains, enabling them to operate in embodied environments, and combining commonsense with encyclopedic knowledge.
This lecture satisfies requirements for CSCI 591: Research Colloquium
Join Zoom Meeting
https://usc.zoom.us/j/97448700889
Meeting ID: 974 4870 0889
Biography: Dr. Filip Ilievski is a Research Scientist in the Center on Knowledge Graphs within the Information Sciences Institute (ISI) at the USC Viterbi School of Engineering. Filip holds a Ph.D. in Natural Language Processing from the Vrije Universiteit (VU) in Amsterdam, where he also worked as a postdoctoral researcher before joining ISI. His research focuses on the role of commonsense and encyclopedic knowledge in filling gaps in human communication. Over the past three years, he mentored around twenty Master's and Ph.D. students, and collaborated with researchers at USC, CMU, Bosch Research, RPI, and the University of Lyon. Dr. Ilievski has over 40 peer-reviewed publications in top-tier venues on commonsense reasoning, information extraction, and knowledge graphs. Dr. Ilievski has also been actively organizing workshops (AAAI'21), tutorials (AAAI'21, ISWC'20, ISWC'21, TheWebConf'22), symposiums (USC), and a special journal issue (Semantic Web Journal) on these topics.
Host: Gale Lucas
Audiences: Everyone Is Invited
Contact: Cherie Carter
-
AME Seminar
Wed, Jan 12, 2022 @ 03:30 PM - 04:30 PM
Aerospace and Mechanical Engineering
Conferences, Lectures, & Seminars
Speaker: Christian Franck, University of Wisconsin-Madison
Talk Title: Cellular NeuroMechanics -“ Concussions, Traumatic Brain Injury and the mysterious Havana Syndrome
Abstract: Current prediction, prevention and diagnosis strategies for mild traumatic brain injuries, including concussions, are still largely in their infancy due to a lack of detailed understanding and resolution of how physical forces give rise to tissue injury at a cellular level. In this talk I will present some recent work on our current understanding of the origin of concussions and traumatic brain injuries and how cells in the brain interpret and react to the physical forces of trauma. Specifically, I will show that the path to a better understanding of traumatic injuries involves addressing a variety of finite deformation, rate-dependent soft matter and cell mechanics problems along the way. Finally, I will provide an update on how our current understanding of the cellular neuromechanics cannot only help shed light on improving our prediction of TBI but also enable us to dissect the physical origin of emerging injuries such as the Havana Syndrome.
Biography: Christian Franck is a mechanical engineer specializing in cellular biomechanics and new experimental mechanics techniques at the micro and nanoscale. He received his B.S. in aerospace engineering from the University of Virginia in 2003, and his M.S. and Ph.D. from the California Institute of Technology in 2004 and 2008. Dr. Franck held a post-doctoral position at Harvard investigating brain and neural trauma. He was an assistant and associate professor in mechanics at Brown University from 2009 - 2018, and is now the Grainger Institute for Engineering Professor in Mechanical Engineering at the University of Wisconsin-Madison.
His lab at the University of Wisconsin-Madison has developed unique three-dimensional full-field imaging capabilities based on multiphoton microscopy and digital volume correlation. Current application areas of these three-dimensional microscopy techniques include understanding the 3D deformation behavior of neurons in the brain during traumatic brain injuries, and the role of non-linear material deformations in soft matter.
He is the acting director of the Center for Traumatic Brain Injury at the University of Wisconsin-Madison and the ONR-funded Physics-based Neutralization of Threats to Human Tissues and Organs (PANTHER) program, which consists of over 24 PIs nationwide. Key objectives of the Panther program are in better detection, prediction, and prevention of traumatic brain injuries by providing accelerated translation from basic science discovery to civilian and warfighter protection solutions.
Host: AME Department
More Info: https://usc.zoom.us/j/93987337017?pwd=MWd2dXBSL1FaR1RPaHNscjJ1NW80UT09
Webcast: https://usc.zoom.us/j/93987337017?pwd=MWd2dXBSL1FaR1RPaHNscjJ1NW80UT09Location: Online event
WebCast Link: https://usc.zoom.us/j/93987337017?pwd=MWd2dXBSL1FaR1RPaHNscjJ1NW80UT09
Audiences: Everyone Is Invited
Contact: Tessa Yao
-
NL Seminar-Translating faster than a keystroke and dumpster diving for training data
Thu, Jan 13, 2022 @ 11:00 AM - 12:00 PM
Information Sciences Institute
Conferences, Lectures, & Seminars
Speaker: Kenneth Heafield, Univ of Edinburgh
Talk Title: Translating faster than a keystroke and dumpster diving for training data
Abstract: REMINDER Meeting hosts only admit guests that they know to the Zoom meeting. Hence, you are highly encouraged to use your USC account to sign into Zoom.
If you are an outside visitor, please inform us at nlg DASH seminar DASH host AT isi DOT edu beforehand so we will be aware of your attendance and let you in.
Machine translation has a deserved reputation for computational cost. But by burning even more GPU time upfront, we can
make inference fast enough to translate thousands of words per second on a desktop or a sentence in under 10 ms on one CPU core. I will talk about optimizations from chopping off transformer heads to writing assembly that make this possible. Software is available at translatelocally.com and coming soon as a Firefox extension. Fast translation was also useful for the ParaCrawl project, where we went dumpster diving on the web for translations and found a few COMET/BLEU points.
Biography: Kenneth Heafield is a Reader/Associate Professor in en-US)
at the University of Edinburgh working on fast and often good machine translation. He coordinates the Bergamot project adding local translation to Firefox, ran the ParaCrawl project, and was friendly competition with ISI in MATERIAL. He wrote kenlm to do large language models before they were cool.
Host: Jon May and Thamme Gowda
More Info: https://nlg.isi.edu/nl-seminar/
Webcast: https://usc.zoom.us/j/93223825200Location: Information Science Institute (ISI) - Virtual Only
WebCast Link: https://usc.zoom.us/j/93223825200
Audiences: Everyone Is Invited
Contact: Pete Zamar
-
The Fred S. Grodins Keynote Lecturer
Fri, Jan 14, 2022 @ 12:30 PM - 01:30 PM
Biomedical Engineering
Conferences, Lectures, & Seminars
Speaker: Cato T. Laurencin, Dr.
Talk Title: Regenerative Engineering of Bone and Other Musculoskeletal Tissues
Abstract: The treatment of injuries to bone that necessitate bone regeneration continues to be a major challenge for the orthopaedic surgeon. This burden is compounded by the constraints of supply and morbidity associated with autograft tissues, the gold standard of repair. The use of allografts, xenografts, or metal and ceramic implants overcomes many of the limitations associated with autografts but fails to provide a viable solution. We have worked in the area of engineering of bone with a focus on biomaterial selection, scaffold development, cell selection, cell material interaction, growth factor delivery, and more recently developing inducible materials.
This entire body of work over more than thirty years has made matrix-based musculoskeletal engineering a viable clinical alternative, and has motivated the establishment of a new field: regenerative engineering. Regenerative Engineering involves new technologies harnessed over the past decade: advanced materials science including nanotechnology, advanced stem cell science, morphogenesis and developmental biology cues, the knowledge and appreciation of physical forces, and clinical translation. Our work has encompassed many aspects of these new technologies and heralds a bright future for the regeneration of bone and other complex tissues.
Biography: Bio: Cato T. Laurencin, M.D., Ph.D. is the University Professor at the University of Connecticut (one of only two at the school). He earned his B.S.E. in Chemical Engineering from Princeton, his Ph.D. in Biochemical Engineering and Biotechnology from M.I.T., and his M.D., Magna Cum Laude, from Harvard Medical School. Dr. Laurencin is a pioneer of Regenerative Engineering. He is an expert in biomaterials science, stem cell technology, nanotechnology and morphogenesis and has worked in the convergence of these areas. The American Institute of Chemical Engineers created the Cato T. Laurencin Regenerative Engineering Founder's Award recognizing his pioneering efforts. In receiving the Spingarn Medal, he was celebrated for his exceptional career that has made him the world's foremost engineer-physician-scientist. The American Association for the Advancement of Science awarded Dr. Laurencin the Philip Hauge Abelson Prize given "for signal contributions to the advancement of science in the United States". He received the National Medal of Technology and Innovation, the nation's highest honor for technological achievement, in ceremonies at the White House. Dr. Laurencin is an elected member of the National Academy of Engineering, the National Academy of Medicine, the National Academy of Sciences, and a fellow of the National Academy of Inventors, the American Academy of Arts and Sciences and the American Association for the Advancement of Science. Renowned internationally, he is a fellow of the National Academy of Sciences India, the Indian National Academy of Engineering, the African Academy of Sciences, The World Academy of Sciences, and is an Academician of the Chinese Academy of Engineering.
Host: BME Professor David D'Argenio
More Info: https://usc.zoom.us/j/98949329502?pwd=YVY5MVU0aFNoa0U1aTZiaFNIdzY0QT09
More Information: Grodins Keynote Lecture 2022 -Dr. Cato T. Laurencin.pdf
Audiences: Everyone Is Invited
Contact: Michele Medina
-
AME Seminar
Wed, Jan 19, 2022 @ 03:30 PM - 04:30 PM
Aerospace and Mechanical Engineering
Conferences, Lectures, & Seminars
Speaker: Ming Tang, Rice University
Talk Title: Elucidating the thermodynamic origins of reaction heterogeneity in lithium-ion batteries
Abstract: During battery cycling, pronounced reaction non-uniformity frequently develops at multiple length scales within electrodes, which adversely impacts battery performance and life by inducing capacity under-utilization, stress concentration and over-(dis)charging. While heterogeneous reactions are typically attributed to mass transport limitations, thermodynamic factors also play an important role and need to be clarified for developing effective mitigation strategies. At the particle level, we reveal how stress could destabilize the lithium (de)lithiation front in single crystalline and polycrystalline intercalation compounds. Stress also provides a fundamental thermodynamic driving force for dendrite growth on lithium metal anodes, which is shown to be effectively suppressed by stress relief. At the cell level, we discover that the reaction distribution within the porous electrode is strongly influenced by how the equilibrium potential of the active material varies with the state of charge. Two types of reaction behavior are identified for common electrode materials, which have significant implications for their applications in thick electrodes. Based on this finding, an analytical model is formulated to provide highly efficient battery performance predictions and optimization in place of traditional battery cell simulations.
Biography: Ming Tang is an Associate Professor in the Department of Materials Science and NanoEngineering at Rice University. After receiving a Ph.D. degree in Materials Science and Engineering from MIT, He worked at Lawrence Livermore National Laboratory as a Lawrence Postdoctoral Fellow and then a staff scientist. In 2013 he joined Shell Oil as a materials and corrosion engineer, and became an assistant professor at Rice University in 2015. His group is currently interested in applying combined modeling and experimental methods to understand mesoscale phenomena in energy storage systems and use the acquired knowledge to guide microstructure design. He is a recipient of the DOE Early Career Award.
Host: AME Department
More Info: https://usc.zoom.us/j/93987337017?pwd=MWd2dXBSL1FaR1RPaHNscjJ1NW80UT09
Webcast: https://usc.zoom.us/j/93987337017?pwd=MWd2dXBSL1FaR1RPaHNscjJ1NW80UT09WebCast Link: https://usc.zoom.us/j/93987337017?pwd=MWd2dXBSL1FaR1RPaHNscjJ1NW80UT09
Audiences: Everyone Is Invited
Contact: Tessa Yao
-
CS Distinguished Lecture: Jean-Baptiste Mouret (Inria) - Quality Diversity Optimization, from robotics to automated design
Tue, Jan 25, 2022 @ 11:00 AM - 12:20 PM
Computer Science
Conferences, Lectures, & Seminars
Speaker: Jean-Baptiste Mouret, Inria
Talk Title: Quality Diversity Optimization, from robotics to automated design
Series: Computer Science Distinguished Lecture Series
Abstract: Diversity algorithms are a new class of optimization algorithms that aim finding a set of high-performing solutions that all behave differently. They are more and more useful in many fields, from automated design, so that designers can use their own expertise to choose the final design, to robotics, so that a robot can quickly and autonomously adapt to unexpected situations. In this talk, I will first introduce MAP-Elites, which is one of the firsts and most popular Quality Diversity algorithms, then I will show a few recent developments, and highlight some of the most promising applications.
Register in advance for this webinar at:
https://usc.zoom.us/webinar/register/WN_Gtk43EXaRc6Q_-8qCfsx5A
After registering, attendees will receive a confirmation email containing information about joining the webinar.
This lecture satisfies requirements for CSCI 591: Research Colloquium.
Biography: Jean-Baptiste Mouret is a senior researcher ("directeur de recherche") at Inria, a French research institute dedicated to computer science and mathematics, in which he conducts researches that intertwine evolutionary algorithms, robotics, and machine learning to make robots more adaptive. Before joining Inria, he was an assistant professor ("maître de conférences) at ISIR (Institute for Intelligent Systems and Robotics), which is part of Université Pierre et Marie Curie - Paris 6 (UPMC, now Sorbonne Université). He obtained a M.S. in computer science from EPITA in 2004, a M.S. in artificial intelligence from the Pierre and Marie Curie University (Paris, France) in 2005, and a Ph.D. in computer science from the same university in 2008. He was the principal investigator of an ERC grant (ResiBots - Robots with animal-like resilience, 2015-2020), which is the most prestigious grant in Europe, and the recipient of a French "ANR young researcher grant (Creadapt - Creative adaptation by Evolution, 2012-2015). His work was featured on the cover of Nature (Cully et al., 2015) and it received the "2017 ISAL Award for Distinguished Young Investigator in the field of Artificial Life, the "Outstanding Paper of 2015 award from the Society for Artificial Life (2016), the French "La Recherche" award (2016), 3 GECCO best paper awards (2011, GDS track; 2017 & 2018, CS track), and the IEEE CEC "best student paper" award (2009).
Host: Stefanos Nikolaidis
Webcast: https://usc.zoom.us/webinar/register/WN_Gtk43EXaRc6Q_-8qCfsx5ALocation: Online - Zoom Webinar
WebCast Link: https://usc.zoom.us/webinar/register/WN_Gtk43EXaRc6Q_-8qCfsx5A
Audiences: Everyone Is Invited
Contact: Computer Science Department
-
Center of Autonomy and AI, Center for Cyber-Physical Systems and the Internet of Things, and Ming Hsieh Institute Seminar Series
Wed, Jan 26, 2022 @ 02:00 PM - 03:00 PM
Ming Hsieh Department of Electrical and Computer Engineering
Conferences, Lectures, & Seminars
Speaker: Dhruva Kartik Mokhasunavisu, Electrical Engineering, University of Southern California
Talk Title: Stochastic Zero-sum Games between Two Competing Teams
Series: Center for Cyber-Physical Systems and Internet of Things
Abstract: Intelligent autonomous systems can be employed in many areas such as transportation, power grids, manufacturing, wildlife conservation etc. Agents in these systems act sequentially over time in uncertain and dynamically evolving environments. In many cases, information about the system is dispersed among the agents and hence, there is asymmetry in the information used by the agents to select their actions. Furthermore, the agents involved may have different (possibly conflicting) objectives. Due to this asymmetry in information and objectives, agents must act in a decentralized and strategic manner. The focus of this talk is on adversarial interactions that can be modeled as stochastic zero-sum games between two teams of agents. In this setting, agents within a team are cooperative. However, the team as a whole is non-cooperative and adversarial with respect to the other team. Finding the value and Nash equilibria of games with incomplete and asymmetric information is notoriously hard. For our game model, we provide a dynamic programming characterization of the value (if it exists). If the value does not exist, the dynamic program provides us with bounds on the upper and lower values. For some specialized game models, we also characterize a min-max strategy for the minimizing team. Further, we propose a computational methodology to solve the dynamic program and illustrate it with the help of an example.
Biography: Dhruva Kartik is a Postdoctoral Researcher at University of Southern California (USC), Los Angeles co-advised by Prof. Rahul Jain and Prof. Pierluigi Nuzzo. He received his Ph.D. in electrical engineering from USC in 2021, where he worked with Prof. Urbashi Mitra and Prof. Ashutosh Nayyar. While he was at USC, he was an Annenberg Fellow from 2015 to 2019. He received his Bachelor of Technology degree in electronics and communication engineering from the Indian Institute of Technology Guwahati (IITG), India, in 2015. His research interests include multi-agent systems, game theory, probabilistic contracts, and reinforcement learning.
Host: Pierluigi Nuzzo, nuzzo@usc.edu
Webcast: https://usc.zoom.us/webinar/register/WN_zyIBh_1gQLmKpMJG0GyLxwLocation: Online
WebCast Link: https://usc.zoom.us/webinar/register/WN_zyIBh_1gQLmKpMJG0GyLxw
Audiences: Everyone Is Invited
Contact: Talyia White
-
AME Seminar
Wed, Jan 26, 2022 @ 03:30 PM - 04:30 PM
Aerospace and Mechanical Engineering
Conferences, Lectures, & Seminars
Speaker: Jacob Leachman, Washington St. University
Talk Title: TBD
Abstract: TBD
Biography: TBD
Host: AME Department
More Info: https://usc.zoom.us/j/93987337017?pwd=MWd2dXBSL1FaR1RPaHNscjJ1NW80UT09
Webcast: https://usc.zoom.us/j/93987337017?pwd=MWd2dXBSL1FaR1RPaHNscjJ1NW80UT09Location: James H. Zumberge Hall Of Science (ZHS) - 252
WebCast Link: https://usc.zoom.us/j/93987337017?pwd=MWd2dXBSL1FaR1RPaHNscjJ1NW80UT09
Audiences: Everyone Is Invited
Contact: Tessa Yao
