| |
|
|
The Department of Computer Science
of Colorado State University, in cooperation with ISTeC
(Information Science and Technology Center), offers the CS
Colloquium series as a service to all who are interested
in computer science. When in-person meetings are possible, most seminars are scheduled for
Monday 11:00AM -- 11:50AM in CSB 130 or Morgan Library
Event Hall. For help finding the locations of our seminar
meetings, consult the on-line CSU campus map. |
Upcoming Events
CS Colloquium Schedule, Fall 2022
|
|
Computer Science Department Colloquium Introduction to the Graduate Program Speaker: Sanjay Rajopadhye, Professor and Graduate Director, Computer Science Department, Colorado State University When: 11:00AM ~ 11:50AM, Monday August 22, 2022  Abstract: Dr. Rajophadye introduces the Computer Science graduate program at CSU. |
|
|
|
Computer Science Department Colloquium CS Faculty Rapid-Fire Presentations of Current Research: Q&A, first session Speaker: Computer Science Faculty, Colorado State University When: 11:00AM ~ 11:50AM, Monday August 29, 2022 Abstract: CS Faculty answers questions about their research, and opportunities in their research group for students at CSU. |
|
|
|
Computer Science Department Colloquium CS Faculty Rapid-Fire Presentations of Current Research: Q&A, second session Speaker: Computer Science Faculty, Colorado State University When: 11:00AM ~ 11:50AM, Monday September 12, 2022 Abstract: CS Faculty answers questions about their research, and opportunities in their research group for students at CSU. |
|
|
|
Computer Science Department Colloquium Pathways to Cyber Education - What Students Need and Employers Want Speaker: Kenneth L Williams, Ph.D. CISSP, Executive Director; Center for Cybersecurity Defense, American Public University When: 11:00AM ~ 11:50AM, Monday September 19, 2022 Abstract: The need for a large cybersecurity workforce has entered our consciousness with echoes in the popular media. This has resulted in huge interest for training and education by individuals seeking to enter the field with various levels of motivation from the need for a greater earning capital to natural curiosity - many of these individuals are departing other career areas and academic pathways seeking career stability and a long-lasting career. While some are carefully choosing to enter the cybersecurity pathway either with guidance and advise from influencers such as high school counselors or parents, many are influenced by social media. Still, some are entering the cybersecurity pathway after first experiencing fulfillment realized through activities such as cyber competitions and other not so honorable pathways such as the Dark Web. On the flip side of this picture are employers that may be considered as either "good" or "bad" actors depending on how they are viewed by society – regardless of this they are the consumers of the individuals that follow the cyber security pathway – seeking to use their skills, knowledge, and abilities. What employers expect of these individuals is the topic of discussion in this talk along with the discovery of what knowledge, skills, and abilities are needed by individuals as an offer to the employer. As a by product this talk will inform as to the optimal educational or training programs to meet the needs of the individuals discussed above. Bio: Dr. Williams is the Executive Director of the Center for Cyber at American Public University, and President of TPA Associates LLC; a small cybersecurity consulting firm. He is a retired US Army IT officer with 24 years of active service and seven years of federal service as a civilian for the US Army. Dr. Williams has a Graduate and a PhD degree in Cybersecurity from Capella University and has focused intently on various aspects of cybersecurity to include; compliance, governance, and other related aspects of cybersecurity mitigation. Dr. Williams served as a subject matter expert in various cybersecurity forums, consults as a cybersecurity SME for various organizations, and authored IT ⁄ cybersecurity curriculums for universities. Dr. Williams has also served as a cybersecurity consulting to various organizations like the US Cybersecurity command, US Coast Guard, and others. He has also served as a panel member on various cybersecurity forums and in the role of featured speaker in others. Dr. Williams holds current certifications in cybersecurity and Information Technology to include the Sec+, Network+, Certified Information System Security Professional (CISSP) and the Information Technology Infrastructure Library (ITIL)v3. He is an accomplished cybersecurity published author and authority with viewpoints in trade journals such as CSO Online and Security Magazine. Dr. Williams has completed the draft for his first cybersecurity book entitled Bridging the Cybersecurity Divide. Dr. Williams resides in Gainesville VA; approximately 40 minutes west of Washington DC. When not working, researching or writing he enjoys spending time with my family, visiting the gym, and riding over long distances with his road bike. |
|
|
|
Computer Science Department Colloquium Programmatic Reinforcement Learning for All Speaker: Dr. Ashutosh Trivedi, University of Colorado Boulder When: 2:00PM ~ 2:50PM, Thursday September 22, 2022 Abstract: Reinforcement Learning (RL) is an optimization-based approach to problem-solving where learning agents rely on scalar reward signals to discover optimal solutions. The recent success of RL has demonstrated its potential as a viable alternative to "human" programming. However, observing these success stories closely, it is evident that significant expertise is required in deploying RL. This expertise is required in designing a suitable approximation architecture and designing a suitable "flat" representation of the environment in a form required by the architecture. Besides, it is also needed to specify objectives in the language of scalar rewards. This rigid interface---in the form of feature constructions, manual approximations, and reward engineering---between the programmers and the RL algorithms is cumbersome and error-prone. The resulting lack of usability and trust contributes towards barriers to entry in this promising field. My group is working towards democratizing reinforcement learning by developing principled methodologies and powerful tools to improve the usability and trustworthiness of RL-based programming at scale. The aforementioned low-level interactions between the programmers and the RL are akin to programming systems in a low-level assembly language. I envision a programmatic approach to RL where the programmers interact with the RL algorithms by writing programs in a high-level programming language expressing the simulation environment, the choices available to the learning agent, and the learning objectives, while an underlying “interpreter” frees the programmer from the burden of feature construction and approximation heuristics demanded by the state-of-the-art RL algorithms. We dub this setting high-level programmatic reinforcement learning (or programmatic RL for short). To realize the promise of improved usability of programmatic RL, we need RL algorithms capable of efficiently handling rich programmatic features (functional recursion and recursive data structures) and complex dynamical models (governed by ordinary differential equations) while guaranteeing convergence to the optimal value. To enable transparent and trustworthy RL, we need translation schemes to compile learning requirements expressed in high-level languages to scalar reward signals. In this talk, I will summarize our efforts and breakthroughs towards a framework for programmatic RL capable of reasoning with formal requirements, real-time constraints, and recursive environments. Bio: Ashutosh Trivedi received his B.Eng. in Computer Science from NIT Nagpur in 2000, his M.Tech. in Electrical Engineering from IIT Bombay in 2003, and his Ph.D. in Computer Science from the University of Warwick in 2009. He was a postdoctoral researcher at the University of Oxford and at the University of Pennsylvania between 2009 and 2012. He was an assistant professor of Computer Science at IIT Bombay from 2013-2015. He joined the University of Colorado Boulder in 2015, where he is currently an assistant professor of Computer Science. He is also a member of the Programming Languages and Verification (CUPLV) Group. His research interests include formal methods, optimization, and game theory with applications in trustworthy AI, cyber-physical systems, software security, and fairness in AI. He is a recipient of an NSF CAREER award, two AFRL fellowships, and a Liverpool-India fellowship. |
|
|
|
Computer Science Department Colloquium Opening the Black Box: Automated Software Analysis for Algorithm Selection Speaker: Lars Kotthoff, Assistant Professor of Computer Science, the University of Wyoming When: 11:00AM ~ 11:50AM, Monday September 26, 2022 Abstract: Impressive performance improvements have been achieved in many areas of AI by meta-algorithmic techniques, such as automated algorithm selection and configuration. However, existing techniques treat the target algorithms they are applied to as black boxes -- nothing is known about their inner workings. This allows meta-algorithmic techniques to be used broadly, but leaves untapped potential performance improvements enabled by information gained from a deeper analysis of the target algorithms. In this talk, we open the black box without sacrificing universal applicability of meta-algorithmic techniques by automatically analyzing algorithms. We show how to use this information to perform algorithm selection, and demonstrate improved performance compared to previous approaches that treat algorithms as black boxes. Bio: Lars Kotthoff is an assistant professor of Computer Science at the University of Wyoming and previously held post-doctoral appointments at the University of British Columbia, Canada, University College Cork, Ireland, and the University of St Andrews, Scotland. His work in meta-algorithmics, automated machine learning, and applying AI to Materials Science has resulted in more than 80 publications with more than 4,200 citations. He is one of the principal developers of the award-winning mlr machine learning software, widely used in academia and industry, and one of the editors of the first book on automated machine learning. |
|
|
|
Computer Science Department Colloquium Closed Loop Perception for Safety-aware Autonomous Systems Speaker: Dr. Deep Samal, School of Computing at the University of Nebraska – Lincoln When: 11:00AM ~ 11:50AM, Monday October 3, 2022 Abstract: Autonomous Systems such as Autonomous Vehicles (AV), robots and drones are being developed for large scale deployments in real world applications such as transportation, agriculture, defense, urban planning etc. To operate safely in such diverse and dynamic scenarios, the perception engine within these systems must be capable of adapting to the dynamic real-time constraints such as latency and energy consumption. This adaptability is not present in the modern perception systems as they are open-loop by design and therefore neither aware nor capable of reacting to the dynamics of a real-world scenario. My research presents the Closed Loop Perception that interprets the perception process in modern autonomous systems as a control system. It creates the notion of 'perception risk' which represents the state of the process by estimating perception failures and then proposes a risk-resource controller that generates feedback signals to dynamically control the resource allocation within the system by using biologically inspired focus-of-attention mechanisms. The proposed Closed Loop Perception System can introspect and adapt to the real-time requirements of an Autonomous System operating in the wild. Bio: Deep Samal is a Post-Doctoral Associate with the School of Computing at the University of Nebraska – Lincoln. He received the M.S. degree in electrical and computer engineering from Georgia Institute of Technology, Atlanta, GA, USA, in 2016, and the Ph.D. degree in electrical and computer engineering from Georgia Institute of Technology, Atlanta, GA, USA, in 2022 under the supervision of Prof. Saibal Mukhopdhyay and Prof. Marilyn Wolf. Before starting his Ph.D., he was with the End-User Computing Team, VMWare, USA. His current research interests include multimodal computer vision, resource efficient autonomous systems and adaptive perception systems. |
|
|
|
Computer Science Department Colloquium Tunneling between Optima: Search at the Edge of Quantum Computing Speaker: Darrell Whitley, Professor, Department of Computer Science, Colorado State University When: 11:00AM ~ 11:50AM, Monday October 10, 2022 Abstract: There is already considerable interest in Quantum AI and Quantum Local Search algorithms. Many quantum optimization tools require that the optimization problem be transformed to QUBO: Quadratic Unconstrained Boolean Optimization. Quantum Computing promises improving moves in constant time and the ability to tunnel between local optima. However, after a problem has been expressed as a QUBO, it is already possible to find improving moves in constant time, and to deterministically tunnel between local minima in O(n) time. In effect, this talk is about what we can already do without Quantum Computing to optimize QUBO problems. This is an important first step before applying Quantum Computing to QUBO problems. Bio: Prof. Darrell Whitley is a faculty member in Computer Science at CSU. He is a former Chair of the Computer Science Department at CSU. He became an ACM Fellow in 2019 and received the IEEE Pioneer In Evolutionary Computation Award in 2022. |
|
|
|
ISTeC Distinguished Lectures FPGA Accelerators in the Cloud Speaker: Dr. Viktor K. Prasanna, Charles Lee Powell Chair in Engineering, Ming Hsieh Department of Electrical & Computer Engineering, Professor of Computer Science, University of Southern California When: 11:00AM ~ 11:50AM, Monday October 17, 2022 Abstract: With recent dramatic advances in Field Programmable Gate Arrays (FPGAs), these devices are being used along with multi-core and novel memory technologies to realize advanced platforms to accelerate complex applications in the Cloud. We will review advances in reconfigurable computing over the past 25 years leading up to accelerators for data science. We will illustrate FPGA-based parallel architectures and algorithms for a variety of data analytics kernels in streaming graph processing and graph machine learning. While demonstrating algorithm-architecture co-design methodology to realize high performance accelerators for graphs and machine learning, we demonstrate the role of modeling and algorithmic optimizations to develop highly efficient Intellectual Property (IP) cores for FPGAs. We show improved performance for two broad classes of graph analytics: iterative graph algorithms with variable workload (e. g., graph traversal, shortest paths, etc.) and machine learning on graphs (e. g., graph embedding). For variable workload iterative graph algorithms, we illustrate dynamic algorithm adaptation to exploit heterogeneity in the architecture. We conclude by identifying opportunities and challenges in exploiting emerging heterogeneous architectures composed of multi-core processors, FPGAs, GPUs and coherent memory. Bio: Viktor K. Prasanna (sites.usc.edu ⁄ prasanna) is Charles Lee Powell Chair in Engineering in the Ming Hsieh Department of Electrical and Computer Engineering and Professor of Computer Science at the University of Southern California. He is the director of the Center for Energy Informatics at USC and leads the FPGA (fpga.usc.edu) and Data Science Labs (dslab.usc.edu). His research interests include parallel and distributed computing, accelerator design, reconfigurable architectures and algorithms and high performance computing. He serves as the Editor-in-Chief of the Journal of Parallel and Distributed Computing. Prasanna was the founding Chair of the IEEE Computer Society Technical Committee on Parallel Processing. He is the Steering Chair of the IEEE International Parallel and Distributed Processing Symposium. He is a Fellow of the IEEE, the ACM and the American Association for Advancement of Science (AAAS). He is a recipient of 2009 Outstanding Engineering Alumnus Award from the Pennsylvania State University and a 2019 Distinguished Alumnus Award from the Indian Institute of Science. He received the 2015 W. Wallace McDowell award from the IEEE Computer Society for his contributions to reconfigurable computing. He is a member of Academia Europaea. |
|
|
|
ISTeC Distinguished Lectures Accelerating Graph Neural Networks Speaker: Dr. Viktor K. Prasanna, Charles Lee Powell Chair in Engineering, Ming Hsieh Department of Electrical & Computer Engineering, Professor of Computer Science, University of Southern California When: 10:00AM ~ 10:50AM, Tuesday October 18, 2022 Abstract: Recently, Graph Neural Networks (GNNs) have been used in many applications leading to improved accuracy and fast approximate solutions. Training as well as Inference in these networks is computationally demanding. Challenges include access to irregular data, large scale sparse as well as dense matrix computations, limited data reuse and heterogeneity in the various stages of the computation. This talk will review our recent work in the Data Science Lab (dslab.usc.edu) and FPGA ⁄ Parallel Computing Lab (fpga.usc.edu) at USC leading up to current trends in accelerators for data science. For graph embedding, we develop GraphSAINT, a novel computationally efficient technique using graph sampling and demonstrate scalable performance. We develop graph processing over partitions (GPOP) methodology to handle large scale graphs on parallel platforms. On a current FPGA device, we demonstrate up to 100X and 30X speed up for full graph GNN computations compared with state-of-the-art implementations on CPU and GPU respectively. We also demonstrate specific accelerators for two widely used GNN models: GraphSAGE and GraphSAINT. We conclude by identifying opportunities and challenges in exploiting emerging heterogeneous architectures towards a general framework for GNN acceleration. Bio: Viktor K. Prasanna (sites.usc.edu ⁄ prasanna) is Charles Lee Powell Chair in Engineering in the Ming Hsieh Department of Electrical and Computer Engineering and Professor of Computer Science at the University of Southern California. He is the director of the Center for Energy Informatics at USC and leads the FPGA (fpga.usc.edu) and Data Science Labs (dslab.usc.edu). His research interests include parallel and distributed computing, accelerator design, reconfigurable architectures and algorithms and high performance computing. He serves as the Editor-in-Chief of the Journal of Parallel and Distributed Computing. Prasanna was the founding Chair of the IEEE Computer Society Technical Committee on Parallel Processing. He is the Steering Chair of the IEEE International Parallel and Distributed Processing Symposium. He is a Fellow of the IEEE, the ACM and the American Association for Advancement of Science (AAAS). He is a recipient of 2009 Outstanding Engineering Alumnus Award from the Pennsylvania State University and a 2019 Distinguished Alumnus Award from the Indian Institute of Science. He received the 2015 W. Wallace McDowell award from the IEEE Computer Society for his contributions to reconfigurable computing. He is a member of Academia Europaea. |
|
|
|
ISTeC Distinguished Lectures Security, Privacy, and Trust Challenges in AI-enabled Cyber-Physical Systems Speaker: Dr. Mani Srivastava, Distinguished Professor, Electrical & Computer Engineering with a joint appointment in Computer Science, University of California, Los Angeles and Amazon Scholar, Amazon When: 11:00AM ~ 11:50AM, Monday October 24, 2022 Abstract: The emerging nexus of Artificial Intelligence (AI) and Cyber-Physical Systems (CPS) is critical to interfacing our digitized society with the analog world it is embedded via sophisticated perception-cognition-communication-action loops. Essential to this vision is the ability to computationally extract rich inferences about the complex events and activities taking place around us; devise actions and interventions that beneficially affect the physical world and nudge human behaviors; and, to do so in a manner that is performant, efficient, and trusted by various stakeholders. Drawing upon experience with the applications of AI-enabled CPS in mobile health, smart environments, and other use domains, the talk will discuss the unique challenges relating to trust, privacy, and security that arise in these systems; describe emerging approaches that address the challenges; and, highlight the importance of considering the socio-technical contexts of these systems. Bio: Mani Srivastava >mbs@ucla.edu< is on the faculty at UCLA where he is a Distinguished Professor in the ECE Department with a joint appointment in the CS Department, and is affiliated with Amazon as an Amazon Scholar. His research is broadly in the area of multimodal sensor information processing, and Human-Cyber-Physical and IoT systems that are learning-enabled, energy-efficient, and secure & trustworthy. His work spans problems across the entire spectrum of architectures, algorithms, and technologies while focusing on applications in mobile health, smart built environments, and military. He is a Fellow of the ACM and the IEEE. More information about his research is available at his lab’s website (http: ⁄ ⁄ www.nesl.ucla.edu) and his Google Scholar profile (https: ⁄ ⁄ scholar.google.com ⁄ citations?user=X2Qs7XYAAAAJ). To arrange a meeting with the speaker, please contact Prof. Anura Jayasumana: anura.jayasumana@colostate.edu. |
|
|
|
ISTeC Distinguished Lectures Neuro-symbolic Architectures for Complex Event Processing in the Internet of Things Speaker: Dr. Mani Srivastava, Distinguished Professor, Electrical & Computer Engineering with a joint appointment in Computer Science, University of California, Los Angeles and Amazon Scholar, Amazon When: 2:00PM ~ 2:50PM, Monday October 24, 2022 Abstract: The combination of deep neural networks (DNNs) with the Internet of Things (IoT) allows sensing and actuation to be performed in our personal, social, and physical spaces in previously unimagined ways. Deep learning methods deployed across the edge-cloud continuum enable IoT systems to make accurate predictions and decisions from high-dimensional and unstructured real-world sensory data while benefiting from the high-performance tensor operations in hardware accelerators. As a result, in many settings, DNNs have entirely replaced symbolic and mechanistic approaches based on algorithms, scientific models, and human knowledge. However, the benefits come with considerably reduced abilities to generalize to new situations, to assure trustworthiness, and to reason about complex spatiotemporal events that require connecting the dots across large spans of time and space. We will present emerging neuro-symbolic approaches that seek to overcome this tension by integrating neural representations with symbolic reasoning. The former allows efficient processing of multimodal sensory inputs to create precepts that assist reasoning and the latter provides interpretability, enforces constraints, allows for human knowledge injection, and acts as regularizers that guide the learning of neural components. The talk will describe the unique capabilities that neuro-symbolic architectures bring to the IoT domain, the research challenges they present, and initial work on neuro-symbolic architectures in practice. Bio: Mani Srivastava >mbs@ucla.edu< is on the faculty at UCLA where he is a Distinguished Professor in the ECE Department with a joint appointment in the CS Department, and is affiliated with Amazon as an Amazon Scholar. His research is broadly in the area of multimodal sensor information processing, and Human-Cyber-Physical and IoT systems that are learning-enabled, energy-efficient, and secure & trustworthy. His work spans problems across the entire spectrum of architectures, algorithms, and technologies while focusing on applications in mobile health, smart built environments, and military. He is a Fellow of the ACM and the IEEE. More information about his research is available at his lab’s website (http: ⁄ ⁄ www.nesl.ucla.edu) and his Google Scholar profile (https: ⁄ ⁄ scholar.google.com ⁄ citations?user=X2Qs7XYAAAAJ). To arrange a meeting with the speaker, please contact Prof. Anura Jayasumana: anura.jayasumana@colostate.edu. |
|
|
|
Computer Science Department Colloquium Trick or Treat: Confessions of a Retired Used Program Salesman or How to Work Less, be more Productive and maybe even have fun doing it. Speaker: Will Tracz, PhD - Lockheed Martin Fellow Emeritus, ACM and IEEE Senior Member, ACM Distinguished Engineer, and semi-professional musician When: 11:00AM ~ 11:50AM, Monday October 31, 2022 Abstract: This talk provides anecdotal insights into the lessons learned from 38 years in Industry, 25 years as a student at university; 6 years as a DARPA PI, 20 years of professional society participation (e.g., Section chair of IEEE, chair of ACM SIGSOFT and ACM SIGMICRO -chair of various computer conferences, and sitting on several industry, academia and professional society boards), and a couple short stints as a professor of Computer Science and Computer Engineering. In addition, this talk will focus on the evolution of software reuse from simple subroutines to objects to open-source software and the evolution of language, machine architecture and operation system support that made it possible. The goal of this talk is to raise awareness of possible professional career opportunities and how to be better prepared for success in industry, academia, or as a professional musician. Bio: Almost nothing is known about this person. Records of his childhood, education, and professional accomplishments were either ill-kept or have been lost in fires of mysterious origins. Here is what we do know. Tracz is a native-born son of American parents and plays several musical instruments, although none very well according to his wife. He took an abnormally long time getting the usual degrees from a couple of East Coast and West Coast institutions of higher learning whose mascots are portrayed as lions, trees, fruits, or knights, in various shades of red, purple, orange, blue, and white. He belongs to the customary societies and organizations and edits several of their publications where he indulges in promoting a revolutionary seamless evolutionary paradigm shift toward the synergistic coordination and management of intellectual effort for the common man on even days of the week. His interests include a run-of-the-mill selection of the current “hot” topics that seem to change based on funding cycles of various program managers. He has some technical publications and music videos, including a book with cartoons in it, so that even managers can get something out of it, and a song, last performed live at Pleasure Island, Disney World Village, Florida to raise money for the "Save the Year 2000 Fund". Finally, he claims an average number of ordinary jobs for experience. He has been a jack hammer operator, professional rock musician and recording artist, college professor, lifeguard, and sit-down comedian. |
|
|
|
ISTeC Distinguished Lectures The Robots Are Coming, The Robots Are Coming: Teaching an Interdisciplinary Course on Robotics+Art Speaker: Dr. Iris Bahar, Department Head and Professor Department of Computer Science Colorado School of Mines When: 11:00AM ~ 11:50AM, Monday November 7, 2022 Abstract: Art, design, computing, and engineering principles are often taught in a siloed fashion. This approach leaves students with a missed opportunity to work together in interdisciplinary teams and learn valuable skills from one another. In my recently taught course, The Robots Are Coming! The Robots Are Coming! we illustrate the power of multidisciplinary study and the beauty of collaboration among students. This course aims to both augment existing artistic robots and design new dynamic interactive creations and encourages exploration of issues regarding spirit, self, technology, language, ethics, and sustainability as starting points for design. Students started the semester elaborating, enhancing, and extending robotic structures donated by artist and co-instructor Eva Goetz with new mechanical, electrical, and software features. As the class rebuilt the existing robots, students gained hands-on understanding of fundamental principles in engineering, computing, design, and collaboration. Students also designed final team projects in the spirit of Eva’s artistic robots that combined design, hardware, and software concepts covered throughout the semester. My talk concludes with some thoughts on the future of STEM education and how courses may be made more inclusive, collaborative, and engaging. Bio: Iris Bahar received the B.S. and M.S. degrees in computer engineering from the University of Illinois, Urbana-Champaign, and the Ph.D. degree in electrical and computer engineering from the University of Colorado, Boulder. She recently joined the faculty at the Colorado School of Mines in January 2022 and serves at Department Head of Computer Science. Before joining Mines, she was on the faculty at Brown University from 1996-2021 and held dual appointments as Professor of Engineering and Professor of Computer Science. Her research interests focus on energy-efficient and reliable computing, from the system level to device level. Most recently this includes the design of robotics systems. She is the 2019 recipient of the Marie R. Pistilli Women in Engineering Achievement Award and the Brown University School of Engineering Award for Excellence in Teaching in Engineering. She is an IEEE fellow and an ACM Distinguished Scientist. To arrange a meeting with the speaker, please contact Prof. Sudeep Pasricha: sudeep.pasricha@colostate.edu. |
|
|
|
ISTeC Distinguished Lectures A Reconfigurable Hardware Library to Enable Real-time Energy-efficient Robot Scene Perception Speaker: Dr. Iris Bahar, Department Head and Professor Department of Computer Science Colorado School of Mines When: 3:00PM ~ 3:50PM, Monday November 7, 2022 Abstract: Perceiving the position and orientation of objects (i.e., pose estimation) is a crucial prerequisite for robots acting within their natural environments. The goal of sampling-based object pose estimation is to infer the probability distribution of the object pose using observed sensor information. In this talk, I will discuss how an approximate implementation of belief propagation, known as Pull Message Passing for Nonparametric Belief Propagation (PMPNBP) can be used to efficiently model and compute the distribution for articulated objects. I will then present a hardware acceleration approach to enable real-time and energy efficient articulated pose estimation. Our approach is on average, 26X more energy efficient than a high-end GPU and 11X faster than an embedded low-power GPU implementation. Moreover, we present a Monte-Carlo Perception Library generated from high-level synthesis to enable reconfigurable hardware designs on FPGA fabrics that are better tuned to user-specified scene, resource, and performance constraints. Bio: Iris Bahar received the B.S. and M.S. degrees in computer engineering from the University of Illinois, Urbana-Champaign, and the Ph.D. degree in electrical and computer engineering from the University of Colorado, Boulder. She recently joined the faculty at the Colorado School of Mines in January 2022 and serves at Department Head of Computer Science. Before joining Mines, she was on the faculty at Brown University from 1996-2021 and held dual appointments as Professor of Engineering and Professor of Computer Science. Her research interests focus on energy-efficient and reliable computing, from the system level to device level. Most recently this includes the design of robotics systems. She is the 2019 recipient of the Marie R. Pistilli Women in Engineering Achievement Award and the Brown University School of Engineering Award for Excellence in Teaching in Engineering. She is an IEEE fellow and an ACM Distinguished Scientist. To arrange a meeting with the speaker, please contact Prof. Sudeep Pasricha: sudeep.pasricha@colostate.edu. |
|
|
|
Computer Science Department Colloquium Parallel Programming with MPI: Successes, Current Challenges, Future Speaker: Anthony Skjellum, PhD, Professor of Computer Science and SimCenter, University of Tennessee at Chattanooga When: 11:00AM ~ 11:50AM, Monday December 5, 2022 Abstract: The rich history and success of explicit parallel programming with message passing plus extensions have supported the growth of multicomputers, clusters, and supercomputers over the past forty plus years, starting with the Caltech Cosmic Cube in 1981. The Message Passing Interface (MPI), a community standard defined 30 years ago, has proven an effective and widely used programming notation (abstractions, syntax, semantics, etc) used pervasively for scale out computing. With a history spanning pre-Terascale through Exascale, MPI has delivered much of the promises it offered originally: portable parallel programs, good performance, and the ability to achieve a degree of performance-portability. In complement, research in portable implementation of MPI has gone on continuously to provide open source products such as MPICH and Open MPI, used widely. MPI's standards are notable for their latitude in how implmentations achieve compliance, and is not a protocol-based standard like TCP ⁄ IP. The Terascale-to-Exascale transformation of scalable architectures (1996-2022) has put stress on the ability to deliver acceptable performance-portability in many cases, deriving from the new complexity of heterogeneous architectures, more so at present than the absolute scale of the number of 'MPI processes' in a scalable execution, or the massive changes in performance of processors, memories, and networks over this span of time. This talk concentrates on explaining the original abstractions, emerging abstractions, and challenges faced for applications and MPI designers to achieve performance portability with accelerator-based systems, changes to node organization, and limitations of abstractions posed in MPI originally. Providing new, revised, and higher-level abstractions, as well as support for modern programming languages, are discussed as ingredients to a solution strategy for MPI as it enters its 4th decade. Bio: Dr. Anthony (Tony) Skjellum studied at Caltech (BS, MS, PhD). His PhD work emphasized portable, parallel software for large-scale dynamic simulation, with a specific emphasis on message-passing systems, parallel nonlinear and linear solvers, and massive parallelism. From 1990-93, he was a computer scientist at LLNL focusing on performance-portable message passing and portable parallel math libraries. From 1993-2003, he was on faculty in Computer Science at Mississippi State University, where his group co-invented the MPICH implementation of the Message Passing Interface (MPI) together with colleagues at Argonne National Laboratory. He has also been active in the MPI Forum since its inception in 1992. From 2003-2013, he was professor and chair at the University of Alabama at Birmingham, Dept. of Computer and Information Sciences. In 2014, he joined Auburn University as Lead Cyber Scientist and led R&D in cyber and High-Performance Computing for over three years. In Summer 2017, he joined the University of Tennessee at Chattanooga as Professor of Computer Science, Chair of Excellence, and Director, SimCenter, where he continues work in HPC (emphasizing MPI, scalable libraries, and heterogeneous computing). He is a co-PI of the DOE ⁄ NNSA PSAAP III Center "Center for Understandable, Performant Exascale Communication Systems" led by the University of New Mexico. He is a senior member of ACM, IEEE, ASEE, and AIChE, and an Associate Member of the American Academy of Forensic Science (AAFS), Digital & Multimedia Sciences Division. |
|
|
|
Computer Science Department Colloquium User-Centred Design of Ubiquitous Computing Technologies: From Interactive Furniture to Farm Robots Speaker: Stacey D. Scott, Ph.D, Professor of Computer Science at the University of Guelph When: 4:00PM ~ 4:50PM, Tuesday December 6, 2022 Abstract: Mark Weiser's now 30-year old vision for ubiquitous computing, where computers would be everywhere and blend into the fabric of our everyday lives, has largely come to pass. Computers may not be as "invisible" as Weiser envisioned, but many everyday objects have become "smart" – or capable of digital computation and ⁄ or control. Computers have also permeated many aspects of our everyday lives, just as he predicted. For over 15 years, Dr. Stacey Scott has designed interfaces for "smart" furniture and spaces, such as interactive tables, walls, and multi-display environments to support many computer-based group activities in home, office, and other domain settings. In a relocation to the University of Guelph in 2016, she began exploring "smart" spaces with other type of inhabitants: farm animals. Leveraging the strong agricultural and animal science roots of UofG, and following a life-long passion for animals and animal welfare, Dr. Scott mounted a new program at the intersection of two emerging fields, "Precision Livestock Farming" and "Animal-Computer Interaction", to explore how "ubiquitous computing" can improve the health and welfare of farm animals. Advances in this space have the potential to improve the economic viability of Canadian farms through the adoption of systems that allow earlier detection and prevention of disease and other livestock health issues. This talk will overview Dr. Scott's past, present, and future "ubiquitous computing" research. Bio: Dr. Stacey Scott is a Professor of Computer Science at the University of Guelph (UofG) and the Director of the Collaborative Systems Laboratory. Prior to joining UofG, she was a faculty member for nine years in Systems Design Engineering at the University of Waterloo. She has had significant impact in the field of surface computing through her work on interactive tabletops and walls and on multi-surface systems that combine personal devices (tablets and smartphones) together with large surfaces to support real-world collaboration. She was awarded the Lasting Impact Award for her contributions to tabletop computing in 2020 from the ACM Conference on Computer Supported Cooperative Work and Social Computing (CSCW). Her general research interests include emerging technology design, human-centred design, animal-centred design, and interface and interaction design. email: stacey DOT scott AT uoguelph DOT ca, web: http: ⁄ ⁄ csl.uwaterloo.ca |
|
|
|
Computer Science Department Colloquium About the Polyhedral Model: Past, Present, and Future Speaker: Patrice Quinton, Professor Emeritus, Ecole normale supérieure de Rennes, IRISA, France When: 3:00PM ~ 3:50PM, Thursday December 8, 2022 Abstract: The Polyhedral Model is a representation of loops as calculations associated to points in integral polyhedra, and a set of mathematical tools allowing such a representation to be used for the generation of parallel algorithms and architectures. During this talk, I will present a subjective point of view, based on my own research trajectory, of the main milestones that made the Polyhedral Model what it is today, and I will present a few directions that research could follow to develop this model and its applications in the future. I will illustrate one of these directions by sketching how architectures for the FFT could be synthesized from their equations, a challenge that was, for a long time, considered to be beyond the scope of the Polyhedral Model. Bio: Patrice Quinton is currently Professor Emeritus of the Ecole normale supérieure de Rennes, and member of the Taran research team of IRISA. https: ⁄ ⁄ www.irisa.fr ⁄ en ⁄ teams ⁄ taran His research concerns mainly the synthesis of parallel accelerators. Until 2015, he was heading the Ecole normale supérieure de Rennes. |
