BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Robotics Institute Carnegie Mellon University - ECPv6.15.12.1//NONSGML v1.0//EN
CALSCALE:GREGORIAN
METHOD:PUBLISH
X-WR-CALNAME:Robotics Institute Carnegie Mellon University
X-ORIGINAL-URL:https://www.ri.cmu.edu
X-WR-CALDESC:Events for Robotics Institute Carnegie Mellon University
REFRESH-INTERVAL;VALUE=DURATION:PT1H
X-Robots-Tag:noindex
X-PUBLISHED-TTL:PT1H
BEGIN:VTIMEZONE
TZID:America/New_York
BEGIN:DAYLIGHT
TZOFFSETFROM:-0500
TZOFFSETTO:-0400
TZNAME:EDT
DTSTART:20250309T070000
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0400
TZOFFSETTO:-0500
TZNAME:EST
DTSTART:20251102T060000
END:STANDARD
BEGIN:DAYLIGHT
TZOFFSETFROM:-0500
TZOFFSETTO:-0400
TZNAME:EDT
DTSTART:20260308T070000
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0400
TZOFFSETTO:-0500
TZNAME:EST
DTSTART:20261101T060000
END:STANDARD
BEGIN:DAYLIGHT
TZOFFSETFROM:-0500
TZOFFSETTO:-0400
TZNAME:EDT
DTSTART:20270314T070000
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0400
TZOFFSETTO:-0500
TZNAME:EST
DTSTART:20271107T060000
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20261106T143000
DTEND;TZID=America/New_York:20261106T153000
DTSTAMP:20261115T000633
CREATED:20260828T170037Z
LAST-MODIFIED:20261113T205734Z
UID:153355-1793975400-1793979000@www.ri.cmu.edu
SUMMARY:Learning in Production: Building Physical AI at Walden Robotics
DESCRIPTION:Abstract: \nWalden Robotics is a full-stack company building both the hardware and physical AI for general-purpose manufacturing robots. Walden was founded in January 2026 and deployed its first robot into production manufacturing two months later. It’s still there and the fleet has been learning on the job ever since. Our core conviction is that the best physical AI will be built via a continuous cycle of deployment and learning grounded in real work. I’ll share lessons from the frontier of end-to-end learning in production manufacturing and why we believe physical AI has reached an inflection point. \nBio: \nBen Burchfiel is the Co-founder and CTO of Walden Robotics\, which develops general-purpose robots that continuously learn and improve while performing real-world industrial tasks. Previously\, at Toyota Research Institute\, Ben co-led a 60+ person robot learning organization responsible for projects including Diffusion Policy\, UMI\, OpenVLA\, and LBM1.0. Ben earned his PhD in computer science at Duke University and completed a postdoctoral fellowship in Brown University’s Department of Computer Science.
URL:https://www.ri.cmu.edu/event/learning-in-production-building-physical-ai-at-walden-robotics/
LOCATION:1403 Tepper School Building
CATEGORIES:RI Seminar,Seminar
ATTACH;FMTTYPE=image/jpeg:https://www.ri.cmu.edu/app/uploads/2026/08/burchfiel.jpeg
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20261105T153000
DTEND;TZID=America/New_York:20261105T163000
DTSTAMP:20261115T000633
CREATED:20261020T160536Z
LAST-MODIFIED:20261020T172927Z
UID:153994-1793892600-1793896200@www.ri.cmu.edu
SUMMARY:Structured Learning for Fast Physical Prediction
DESCRIPTION:Abstract:\nPhysical simulation is essential for robotics and engineering design\, but remains computationally expensive\, especially for complex geometries and long time horizons. Learning-based approaches offer faster predictions\, but often lack the mathematical structure needed for reliable simulation.\nIn this talk\, I will present three approaches that combine learning with mathematical structure: continuous reduced-order representations for generalization across shapes\, geometric lifting for handling discontinuities\, and Koopman-based linear operators for efficient long-horizon prediction. Together\, these methods demonstrate how structured learning can enable faster\, more expressive\, and more reliable physical simulation.\n\nBiography:\nChang Yue is a Ph.D. student in Computer Science at the University of Toronto\, advised by Prof. Eitan Grinspun. Her research focuses on mathematically structured learning for fast and generalizable physical simulation. Her work has received the SIGGRAPH 2025 Best Paper Award and Best Paper Award Honorable Mention. She was also recognized as a WiGRAPH Rising Star in 2026.\n\nHomepage: https://changy1506.github.io/\n\nSponsor:\nThe VASC seminar is generously sponsored by HeyGen\, an all-in-one\nAI-powered video generation platform that leverages advances in\ncomputer vision\, generative modeling\, and multimodal learning to make\nhigh-quality video creation both scalable and accessible.
URL:https://www.ri.cmu.edu/event/structured-learning-for-fast-physical-prediction/
LOCATION:3305 Newell-Simon Hall
CATEGORIES:Seminar,VASC Seminar
ATTACH;FMTTYPE=image/jpeg:https://www.ri.cmu.edu/app/uploads/2026/10/11-5-26.jpg
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20261102T123000
DTEND;TZID=America/New_York:20261102T140000
DTSTAMP:20261115T000633
CREATED:20261027T170243Z
LAST-MODIFIED:20261027T170243Z
UID:154050-1793622600-1793628000@www.ri.cmu.edu
SUMMARY:RI PhD Thesis Proposal - Wenli Xiao
DESCRIPTION:Date: November 2\, 2026\nTime: 12:30–2:00 PM EST\nType: RI PhD Thesis Proposal\nZoom Link\nLocation: NSH 4305\nTitle: Scalable Robot Self-Improvement through Real-World Interaction\nWho: Wenli Xiao \n \nAbstract: \nA central ambition of AGI is to build agents that can continually acquire new capabilities from their own experience. For embodied agents\, this means learning through actions whose consequences are tested in the physical world: identifying gaps in their capabilities\, seeking useful experience\, and using the resulting feedback to improve. This thesis investigates how real-world interaction can support such self-improvement\, and how the learning process itself can become increasingly autonomous and scalable. \nI study two complementary mechanisms for learning from experience—in-context learning and post-training—and the role of agents in directing that learning. The research develops through four parts: \n\nLearning from Experience in Context: WoCoCo studies whole-body control conditioned on contact and task goals\, providing a precursor in context-conditioned behavior. AnyCar investigates how a dynamics model can use recent state–action history to adapt its predictions across vehicles and changing conditions. This establishes a concrete form of learning from interaction during deployment.\nTurning Experience into Lasting Capability: PLD uses residual reinforcement learning to develop specialist corrections and distills their experience into a generalist policy. ASAP uses physical feedback to align simulation and improve policies through further training. These works investigate how experience can produce improvements that persist beyond the current context.\nAgentic Self-Improvement: ENPIRE enables coding agents to manage physical experiments\, evaluate outcomes\, and iterate on policy improvement. It takes a step toward systems that organize their own learning\, with agents selecting and executing experiments to improve robot behavior.\nScaling Self-Improvement (Proposed Work): I propose a unified autoregressive robotic model that integrates reasoning\, action generation\, and in-context learning\, aiming for a base model that can interpret feedback and be steered toward new goals.\n\nThe proposed work will test whether these ingredients support sustained improvement over repeated learning cycles\, retention of acquired capabilities\, and transfer to new conditions with less human intervention. The broader goal is embodied intelligence whose capabilities grow through an increasingly self-directed cycle of interaction\, learning\, and discovery. \nCOMMITTEE MEMBERS \nGuanya Shi (Chair)\nChangliu Liu\nAviral Kumar\nSergey Levine\, UC Berkeley
URL:https://www.ri.cmu.edu/event/ri-phd-thesis-proposal-wenli-xiao/
LOCATION:Newell-Simon Hall 4305
CATEGORIES:PhD Thesis Proposal,Student Talks
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20261030T143000
DTEND;TZID=America/New_York:20261030T153000
DTSTAMP:20261115T000633
CREATED:20260828T165840Z
LAST-MODIFIED:20261031T152920Z
UID:153351-1793370600-1793374200@www.ri.cmu.edu
SUMMARY:Expanding What Robots Can Learn in the Real World
DESCRIPTION:Abstract: Two robots can be equally capable today yet very different to teach tomorrow. As foundation models expand what robots can already do\, an equally important question is how to expand their learning potential: the capabilities they can acquire next with the limited experience and supervision available to them. In this talk\, I will argue that expanding this potential requires asking not only how robots learn\, but what they should learn and reuse to acquire new capabilities safely and efficiently through real-world interaction. I will present selected learning systems for mobile manipulation that address this question in three ways: learning reusable action spaces that support safer and more efficient physical exploration; using human demonstrations and corrections to guide real-world refinement; and selecting or collecting experience according to its contribution to policy improvement. I will close with a perspective on memory and continual learning as steps toward making these improvements persist across tasks and deployments. Together\, these directions aim to expand the learning frontier by preserving useful knowledge while enabling the changes that new tasks require. \n\n\nBio: Roberto Martín-Martín is an Assistant Professor of Computer Science at the University of Texas at Austin\, where he directs the Robot Interactive Intelligence Lab (RobIn). He builds robotic systems that learn in the real world\, studying how robots can acquire useful skills from people and physical interaction\, learn safely from limited experience\, and keep improving continuously over time. He pursues these questions through mobile and whole-body manipulation\, contact-rich interaction\, and long-horizon tasks in everyday environments. He earned his Ph.D. at TU Berlin under Oliver Brock and conducted postdoctoral research at the Stanford Vision and Learning Lab with Fei-Fei Li and Silvio Savarese. His work has received the RSS Best Systems Paper Award\, ICRA Best Paper Award\, and IROS Best Mechanism Award\, and was awarded the NSF CAREER Award\, Amazon Faculty Award\, Meta Research Award and TRI Young Faculty Researcher Award. He also serves the robotics community as Chair of the IEEE Technical Committee on Mobile Manipulation and as a co-founder of QueerInRobotics.
URL:https://www.ri.cmu.edu/event/expanding-what-robots-can-learn-in-the-real-world/
LOCATION:1403 Tepper School Building
CATEGORIES:RI Seminar,Seminar
ATTACH;FMTTYPE=image/jpeg:https://www.ri.cmu.edu/app/uploads/2026/08/rmm-scaled.jpeg
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20261028T110000
DTEND;TZID=America/New_York:20261028T123000
DTSTAMP:20261115T000633
CREATED:20261021T133556Z
LAST-MODIFIED:20261021T133556Z
UID:154008-1793185200-1793190600@www.ri.cmu.edu
SUMMARY:RI PhD Thesis Proposal - Chen Wu
DESCRIPTION:Date: October 28\, 2026\nTime: 11:00 AM -12:30 PM ET\nType:  RI PhD Thesis Proposal\nZoom Link\nLocation: NSH 3305\nTitle: Learning to Explore and Learning from Exploration for Language Models\nWho: Chen Wu \nAbstract: \n\nLearning from experience offers a way for language models to improve beyond the solutions available in their training data. Instead of relying only on demonstrations\, a model can attempt problems\, receive feedback\, and learn from the outcomes. This process takes many forms\, such as reinforcement learning with verifiable rewards\, search at test time\, training via self-improvement\, and the integration of knowledge that arrives after training. In all of these scenarios\, two capabilities matter: (1) exploring a broad range of solutions\, and (2) learning from exploration to improve future performance. In practice\, both still remain challenging — models often repeat similar attempts even when they fail\, and learning from their outcomes is difficult when feedback is unreliable or when updates degrade existing capabilities. This thesis develops methods to address both challenges. \nFor learning to explore\, we study why language models generate a narrow range of solutions and how to encourage broader exploration. We first examine the role of training objectives in simple algorithmic tasks\, finding that next-token prediction is myopic\, while multi-token objectives and input-side randomness produce more diverse and original outputs. We then study the effect of post-training\, which concentrates a model’s reasoning into a small number of modes. We introduce mode-conditioning (ModC)\, which makes these modes controllable and allocates test-time samples across them. \nFor learning from exploration\, we study how models can provide useful feedback to themselves when exploring and can incorporate new experience without losing existing capabilities. We first introduce self-trained verification (STV)\, which teaches a model to identify errors in its own solutions. The resulting verifier improves reasoning through iterative refinement\, especially on hard problems. We then address the problem of forgetting when learning from new experience. We show that the degradation caused by off-policy training is not inherent to the data\, but depends on how the update is applied. We therefore introduce grafting\, which learns the update on an earlier checkpoint and transfers it to the post-trained model while masking the directions to which the post-trained model is most sensitive. Grafting outperforms more complex and expensive baselines on both new and old tasks. \nBuilding on these results\, the proposed work continues under the two directions. For learning to explore\, I will study how to guide exploration on problems beyond a model’s current ability\, with hints\, reference solutions\, and verifier feedback\, without the model depending on or exploiting the guidance. For learning from exploration\, I will study how new experience can be integrated with pretrained knowledge so that the two compose\, and how a model can learn more from less experience by learning how to learn. \nCOMMITTEE MEMBERS\n\nAditi Raghunathan\, Chair\nJeff Schneider\nMax Simchowitz\nRuslan Salakhutdinov\nYejin Choi\, Stanford University\n\nLink to thesis proposal
URL:https://www.ri.cmu.edu/event/ri-phd-thesis-proposal-chen-wu/
CATEGORIES:PhD Thesis Proposal,Student Talks
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20261027T140000
DTEND;TZID=America/New_York:20261027T153000
DTSTAMP:20261115T000633
CREATED:20261019T132814Z
LAST-MODIFIED:20261021T141920Z
UID:153979-1793109600-1793115000@www.ri.cmu.edu
SUMMARY:RI PhD Thesis Proposal - Sriram Narayanan
DESCRIPTION:Who: Sriram Narayanan\nType: RI Thesis Proposal\nDate: Tuesday\, October 27\, 2026\nTime: 02:00PM – 03:30PM (ET)\nLocation: NSH 4305 \nZoom link: Here \nTitle: Thermal Computational Imaging through Light and Heat Transport \nAbstract: \nComputer vision seeks to understand the physical world from images\, yet much of this understanding remains rooted in the appearance of reflected visible light. Objects also absorb and emit radiation\, exchange heat with their surroundings\, and change their temperature over space and time. These processes encode information about materials\, geometry\, and physical interactions that conventional visible cameras often cannot access. Thermal imaging makes aspects of this energy exchange observable\, opening new possibilities for computational imaging. \nThis dissertation develops a physics-based framework for thermal computational imaging by treating light and heat transport as complementary sources of scene information.\nObservations across wavelength and time depend differently on geometry\, materials\, illumination\, and temperature. Connecting these observations through physical models helps distinguish scene properties that remain ambiguous in individual measurements. \nWe develop this premise through a sequence of inverse problems. Spectral and temporal thermal observations enable separation of radiation emitted by an object from radiation reflected from its surroundings\, supporting temperature recovery even when material emissivity is unknown. Next\, we show how the thermal consequences of absorbed light complement visible reflection\, allowing surface reflectance and shading to be separated without explicitly recovering geometry or lighting. We then establish heat conduction as a geometric cue\, recovering three-dimensional shape from the spatial relationships governing temperature evolution across dark\, textured\, and visibly transparent objects. Combining conduction with shading further allows local shape ambiguities to be resolved through their complementary geometric constraints. \nThe proposed work extends these connections into thermal simulation within VisionSIM\, linking scene illumination\, transient heat transport\, emitted and reflected radiation\, and sensor dynamics to predict thermal observations. Together\, the completed and proposed work connect scene reconstruction and image synthesis through the coupled physics of light and heat. They establish energy exchange as both a source of constraints for recovering scene properties and a basis for predicting thermal observations. \nThesis Committee Members:\nSrinivasa Narasimhan (Chair)\nAswin Sankaranarayanan\nIoannis Gkioulekas\nRavi Ramamoorthi (UCSD)\nShree Nayar (Columbia) \n\nDraft of the Thesis Proposal Document
URL:https://www.ri.cmu.edu/event/ri-phd-thesis-proposal-sriram-narayanan/
LOCATION:Newell-Simon Hall 4305
CATEGORIES:PhD Thesis Proposal,Student Talks
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20261026T153000
DTEND;TZID=America/New_York:20261026T163000
DTSTAMP:20261115T000633
CREATED:20261020T150818Z
LAST-MODIFIED:20261020T150818Z
UID:153989-1793028600-1793032200@www.ri.cmu.edu
SUMMARY:Controllable\, Efficient and Personalized Generative AI: From Long-Horizon Video to Interactive World Models
DESCRIPTION:Abstract:  Generative models can produce compelling images and videos\, yet building systems that people can reliably direct and personalize remains a fundamental challenge. These systems must support new forms of control\, generate over extended horizons within practical computational budgets\, and adapt to individual concepts and preferences. In this talk\, I will explore these challenges through a central question: How can we make visual generation controllable\, efficient\, and personal?\nI will present three complementary directions toward this goal. First\, Infinity-RoPE (CVPR’26) revisits how autoregressive video models represent temporal context\, unlocking long-horizon generation and responsive control through inference-time interventions. Next\, VideoMLA (NeurIPS’26) addresses the computational demands of extended generation by learning compact representations of past context\, substantially reducing memory requirements while preserving generation quality. Finally\, SPAWN (NeurIPS’26) introduces personalized concepts into interactive world models\, allowing users to bring their own subjects into generated environments as they navigate them. \nI will close with broader directions for video generation\, including new distillation methods for efficient generation\, improving physical plausibility (HiPhy\, NeurIPS’26)\, and aligning generation with desired outcomes through reinforcement learning (DPP-GRPO\, CVPR’26). This research advances a broader vision for generative cinematography\, supported by an NSF CAREER’26 award\, toward generative worlds that remember the past\, respond to our choices\, and evolve as we interact with them.\n\n\nBio:  Pinar Yanardag is an Assistant Professor of Computer Science at Virginia Tech\, where she leads GEMLAB (Generative Modeling Lab). Her research focuses on controllable\, efficient\, and personalized generative AI\, with an emphasis on video generation and interactive world models. She earned her PhD from Purdue as a Fulbright Fellow and was a postdoctoral researcher at the MIT Media Lab\, where her generative AI projects attracted over one million participants and media coverage from TIME\, The Washington Post\, and CNN. She received an NSF CAREER award for her research on generative cinematography and was recognized as an Outstanding Area Chair at CVPR 2026. Her research is supported by industry partners including Google\, Adobe\, Snap\, and Qualcomm.\n\n\nHomepage: https://pinguar.org\n\n\nSponsor:\nThe VASC seminar is generously sponsored by HeyGen\, an all-in-one\nAI-powered video generation platform that leverages advances in\ncomputer vision\, generative modeling\, and multimodal learning to make\nhigh-quality video creation both scalable and accessible.
URL:https://www.ri.cmu.edu/event/controllable-efficient-and-personalized-generative-ai-from-long-horizon-video-to-interactive-world-models/
LOCATION:Newell-Simon Hall 3305
CATEGORIES:Seminar,VASC Seminar
ATTACH;FMTTYPE=image/jpeg:https://www.ri.cmu.edu/app/uploads/2026/10/10-26-26-pinaryanardag_biopic.jpg
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20261026T100000
DTEND;TZID=America/New_York:20261026T113000
DTSTAMP:20261115T000633
CREATED:20261020T200109Z
LAST-MODIFIED:20261021T133756Z
UID:154003-1793008800-1793014200@www.ri.cmu.edu
SUMMARY:RI PhD Thesis Proposal - Yilin Wu
DESCRIPTION:Date: October 26\, 2026\nTime: 10:00-11:30 AM ET\nType:  RI PhD Thesis Proposal\nZoom Link\nLocation: NSH 4305\nTitle: Reconciling Adaptation and Generality in Robot Policies\n\nWho: Yilin Wu \nAbstract: \n\nGeneralist robot policies—such as vision-language-action models (VLAs)—are trained end-to-end on large observation-action datasets and can perform increasingly diverse manipulation tasks. Yet these policies remain unreliable amid the complexity of real environments and user contexts. A common remedy is to finetune the entire policy for a particular deployment\, but this may degrade the pretrained capabilities that make generalist policies valuable. This raises the central question of this thesis: how should robot policies be adapted to particular users and environments while preserving their generality?\n\nThis thesis argues that maximizing adaptation while minimizing loss of generality requires identifying what should be adapted and choosing how to adapt it in the least invasive way. For example\, a deployment may call for changes to the task specification (what the user wants done)\, the state representation (which aspects of the scene matter for the task)\, or the behavioral support (the range of actions the policy can produce). At each of these levels\, the desired behavior may already lie within the policy’s pretrained capabilities and only need to be elicited\, or the policy may need to acquire new knowledge or skills through additional training.\n\nWe study two corresponding adaptation mechanisms: inference-time steering\, which guides pretrained action distribution toward the target while keeping policy parameters fixed\, and targeted training\, which learns additional context or behaviors through limited parameter updates. For steering\, we first propose language-based verification methods that check generated actions against user instructions and reasoning plans. Next\, to handle incomplete state information in contact-rich manipulation\, we introduce a bi-level visual and tactile steering algorithm. Third\, to steer the action distribution in open-world settings\, we contribute a world model that predicts the consequences of candidate behaviors while balancing fidelity\, temporal consistency\, and inference efficiency.\n\nSteering\, however\, can only elicit behaviors the pretrained policy already supports. We turn to targeted training methods with parameter-efficient updates to improve the base policy via learning a compact multi-stream memory that conditions the policy on history-dependent state\, and residual policies that incorporate user corrections to expand the policy’s behavioral capabilities. Finally\, the proposed work revisits the role of meta-learning for adapting generalist robot policies. We propose separating shared representation learning across tasks from task-specific action learning using limited demonstrations. The goal is adaptation with fewer demonstrations and trainable parameters\, with explicit retention of pretrained capabilities\nCOMMITTEE MEMBERS\nAndrea Bajcsy (Chair)\nGuanya Shi\nOliver Kroemer\nAnirudha Majumdar\,  Princeton University
URL:https://www.ri.cmu.edu/event/ri-phd-thesis-proposal-yilin-wu/
LOCATION:Newell-Simon Hall 4305
CATEGORIES:PhD Thesis Proposal,Student Talks
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20261023T143000
DTEND;TZID=America/New_York:20261023T153000
DTSTAMP:20261115T000633
CREATED:20260828T165634Z
LAST-MODIFIED:20261023T223452Z
UID:153347-1792765800-1792769400@www.ri.cmu.edu
SUMMARY:Simulation-Driven Robot Learning for the Lazy Researcher
DESCRIPTION:Abstract: This talk will delve into the workflow of building robot learning methods powered by large-scale physics simulation. While the promise of simulation is low-effort data generation that scales cheaply with human effort\, the reality is often not as simple\, with considerable human effort required to actually make simulation a useful tool for real-world robotic manipulation. In this talk\, I will study how to develop methods that explicitly reduce manual human effort in enabling simulation as a data generation tool – from behavior generation\, to model selection\, to enabling downstream transfer to the real-world. In doing so\, I will outline an effort-efficient path to leveraging simulation as a data source for robot learning – hopefully to make robotics easier for lazy researchers like me!\n\nBio: Abhishek Gupta is an assistant professor in computer science and engineering at the Paul G. Allen School of Computer Science and Engineering at the University of Washington since 2022. He leads the Washington Embodied Intelligence and Robotics Development lab focusing on robot learning and reinforcement learning. Previously\, he was a postdoctoral scholar at MIT\, collaborating with Russ Tedrake and Pulkit Agrawal. Prior to that he received his Ph.D. and B.S degrees from UC Berkeley\, working with Sergey Levine and Pieter Abbeel. Abhishek is the recipient of IEEE RAS Early Career Award\, Toyota Research Institute Young Investigator award and an Amazon Science Hub award\, along with award nominations at several top conferences and workshops. His research interests lie in scalable reinforcement learning methods for robot learning\, in particular methods for continual adaptation in the real world and what role simulation can play in this process.
URL:https://www.ri.cmu.edu/event/simulation-driven-robot-learning-for-the-lazy-researcher/
LOCATION:1403 Tepper School Building
CATEGORIES:RI Seminar,Seminar
ATTACH;FMTTYPE=image/jpeg:https://www.ri.cmu.edu/app/uploads/2026/08/gupta.jpeg
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20261021T130000
DTEND;TZID=America/New_York:20261021T140000
DTSTAMP:20261115T000633
CREATED:20261021T142919Z
LAST-MODIFIED:20261021T142919Z
UID:154012-1792587600-1792591200@www.ri.cmu.edu
SUMMARY:RI PhD Speaking Qualifier - Gabrielle Ohlson
DESCRIPTION:Date: Wednesday\, October 21\nTime: 1 – 2 pm\nRoom location: GHC 6115\nZoom link: https://cmu.zoom.us/j/97081285448?pwd=OvaBlpZG3whhrdtg3QSUP0bGPipbKW.1\n\nTitle: Designing & Evaluating Machine-Knitted Stim Objects for Autistic Adults\n\nAbstract:\nStimming is a valuable coping strategy and form of expression for Autistic people. However\, research has historically scrutinized stimming through a negative\, corrective model. When research instead prioritizes the Autistic perspective\, an avenue opens for custom stimming resources that allows Autistic people to experience meaningful and stimulating interactions with sensory materials. Stimming can involve purpose-made objects (“stim objects”)—often small and portable\, these objects come in a variety of textures\, colors\, functions\, and sensory elements. Most commercial stim objects are mass-produced using traditional polymers\, with little focus on how the material meets the user’s sensory needs. Fabrics are ubiquitous tactile materials with intricate\, tunable properties. In my talk\, I will show that knitted fabrics make highly effective stim objects for this reason—designers can tune their material properties at the micro stitch-level\, allowing for parameterized textural and interactive possibilities that are well suited for manipulation by hand.\n\nMy talk will cover my multi-stage research process to answer the question: How can we design custom machine-knitted stim objects that meet the diverse sensory needs of Autistic adults? I will begin by covering the background survey I conducted to gather data to inform the subsequent design and fabrication stage\, followed by an overview of my follow-up user-study for evaluating prototypes with feedback from Autistic adults. Finally\, I will discuss future work with applications for custom stim objects\, improved commercial resources\, and design guidelines based on relevant sensory and material axes.\n\nCommittee members:\nJim McCann (co-advisor)\nAndrew Begel (co-advisor)\nJodi Forlizzi\nYuichi Hirose
URL:https://www.ri.cmu.edu/event/ri-phd-speaking-qualifier-gabrielle-ohlson/
CATEGORIES:PhD Speaking Qualifier,Student Talks
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20261020T153000
DTEND;TZID=America/New_York:20261020T170000
DTSTAMP:20261115T000633
CREATED:20261012T135942Z
LAST-MODIFIED:20261012T135942Z
UID:153954-1792510200-1792515600@www.ri.cmu.edu
SUMMARY:RI PhD Thesis Defense - Ruixuan Liu
DESCRIPTION:Date: October 20\, 2026\nTime: 3:30–5:30 PM ET\nType:  RI PhD Thesis Defense\nZoom Link\nLocation:GHC 4405\nTitle: Prompt-to-Product: Towards Robots that Imagine\, Reason\, and Build\nWho: Ruixuan Liu \nAbstract:\nRecent advances in artificial intelligence (AI) and robotics have opened new possibilities for helping people turn ideas into physical assemblies. Large language models\, multimodal foundation models\, and other generative models can interpret user intent and generate designs\, while robots can perform increasingly complex assembly tasks. Yet these capabilities often remain disconnected. A design that satisfies a user’s request may be physically unstable or beyond a robot’s ability to build. Meanwhile\, a robot proficient at individual manipulation tasks may still depend on detailed\, step-by-step commands. This thesis pursues the vision of PROMPT-TO-PRODUCT: robots that move beyond prescribed actions to help people imagine new assemblies\, reason about their feasibility\, and build them. Using brick assembly as a platform\, it investigates the complete process from abstract user intent to physical construction. Its central argument is that reasoning about physics\, assembly properties\, and robot capabilities must bridge design generation and robotic execution\, so that physical feasibility informs both the imagined assembly and the process of building it. \n\nThe first part of this thesis develops methods for reasoning about the physical properties of brick assemblies\, laying the foundation for the design and construction methods that follow. We present STABLELEGO\, a computational tool for analyzing structural stability\, and a full-stack physics simulator that extends stability analysis to high-fidelity simulation of brick motions and interactions. The second part investigates how abstract user ideas can be translated into concrete assembly designs. We introduce BRICKGPT\, which generates stable and buildable interlocking brick assemblies from text prompts. We then develop an agentic framework for zero-shot assembly design from text and image inputs\, which moves beyond BRICKGPT’s restricted component library to support diverse components and articulated structures. The third part develops the manipulation and coordination capabilities needed for precise\, long-horizon assembly. We present an end-of-arm tool that manipulates small bricks with high precision\, and APEX-MR\, an asynchronous coordination algorithm that enables multiple robot arms to build complex structures collaboratively. We then develop visuomotor policies through sim-to-real learning\, enabling precise assembly with general-purpose grippers on lightweight robots and extending these capabilities beyond industrial platforms. The fourth part integrates reasoning\, generation\, and manipulation into the complete PROMPT-TO-PRODUCT system. It connects STABLELEGO\, BRICKGPT\, and the dual-arm construction system BRICKMATIC to transform natural-language descriptions into physical assemblies\, with stability reasoning guiding both design generation and construction. End-to-end demonstrations and a user study evaluate the full workflow.\n\nTogether\, these contributions connect expressive design generation with the physical reasoning and manipulation required for construction. They expand the range of assemblies that can be designed\, the physical behaviors that can be simulated\, and the robotic platforms that can build them. Ultimately\, this thesis advances the vision of robots that imagine\, reason\, and build alongside people\, helping users realize their ideas without specifying every step of design and construction. \n\n\n\n\nCOMMITTEE MEMBERS:\nChangliu Liu (Chair) \n\nOliver Kroemer\nJean Oh\nJun-Yan Zhu\nKen Goldberg\, UC Berkeley
URL:https://www.ri.cmu.edu/event/ri-phd-thesis-defense-ruixuan-liu/
LOCATION:Gates Hillman Center 4405
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20261019T153000
DTEND;TZID=America/New_York:20261019T163000
DTSTAMP:20261115T000633
CREATED:20261009T175301Z
LAST-MODIFIED:20261009T175301Z
UID:153944-1792423800-1792427400@www.ri.cmu.edu
SUMMARY:“Lawless Progress on Wicked Problems”: Shifting the Roles of Humans in Human-AI Systems
DESCRIPTION:Abstract:   AI is rapidly changing the roles that humans play as developers and users of new technology. In this talk\, I will discuss three of those shifting roles and how my group addresses them through technical innovation and interdisciplinary design. Across these projects\, understanding human roles is central to determining what systems should optimize\, what data they need\, and how they can be trusted.\nHumans as objective functions: Many real-world applications involve optimizing user-specific objectives\, but users’ descriptions of those objectives can be vague\, inconsistent\, or even self-contradictory. I will discuss our work on Interactive Explainable Ranking (Best Paper\, CHI 2026)\, which combines optimization\, AI\, and interaction design to help users identify and resolve contradictions in how they evaluate different items—from open-ended course projects to purchasing options or contestants in a competition to find the best cat. \nHumans as a source of data: AI is only effective if we have the right data\, and many important applications depend on specific or individualized data that needs to be captured in uncontrolled settings. Multiple projects in my group focus on designing systems that help humans capture such data using mobile devices.  This work connects capture guidance\, sampling theory\, computational imaging\, and interaction design to support downstream learning\, with applications spanning remote healthcare\, infrastructure\, and agriculture. \nHumans as adversaries: Some advances in AI also carry significant potential for harm\, which calls for designing new approaches to protect against adversarial use. One problem that AI has made particularly challenging is video forensics. I will discuss my group’s work on designing coded illumination systems to add near-imperceptible watermarks to lighting in high-stakes environments. These watermarks make it easier to detect manipulation in recorded videos without requiring access to or control over recording cameras. \nBio:  Abe Davis is an assistant professor in the Computer Science Department at Cornell University\, where his group works at the intersection of human-AI systems and visual computing. He earned his Ph.D. in EECS at MIT. His dissertation received the Sprowls Award for Best PhD Thesis in Computer Science at MIT\, as well as the ACM SIGGRAPH Outstanding Doctoral Dissertation Honorable Mention Award. He has received the NSF CAREER Award and a Sloan Research Fellowship\, and has published award-winning papers in venues across visual computing and HCI. He has also been recognized in Forbes’ “30 Under 30” and Business Insider’s “50 Scientists Who Are Changing the World.” \nHomepage: Abe Davis’s Group \nSponsor:\nThe VASC seminar is generously sponsored by HeyGen\, an all-in-one\nAI-powered video generation platform that leverages advances in\ncomputer vision\, generative modeling\, and multimodal learning to make\nhigh-quality video creation both scalable and accessible.
URL:https://www.ri.cmu.edu/event/lawless-progress-on-wicked-problems-shifting-the-roles-of-humans-in-human-ai-systems/
LOCATION:Newell-Simon Hall 3305
CATEGORIES:Seminar,VASC Seminar
ATTACH;FMTTYPE=image/jpeg:https://www.ri.cmu.edu/app/uploads/2026/10/10-19-26-AbeHeadshotFormalish.jpg
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20261016T120000
DTEND;TZID=America/New_York:20261016T130000
DTSTAMP:20261115T000633
CREATED:20261011T235146Z
LAST-MODIFIED:20261011T235444Z
UID:153949-1792152000-1792155600@www.ri.cmu.edu
SUMMARY:Global Optimization for Robotics in the Age of Learning
DESCRIPTION:Abstract:\nOptimization is the bedrock of robotics\, and robotics is a great stress\ntest for optimization methods. High dimensionality\, nonconvexity\, and\nnonsmoothness make many robotics problems notoriously hard to solve\,\nwhether with learning-based or model-based strategies.\nIn this talk\, I will discuss our recent advances in bringing\nlearning-based techniques together with optimization principles to\naddress this. First\, I will present extensions of moment-based\noptimization to data-driven\, nonparametric problems that retain partial\nglobal optimality guarantees. I will discuss how these methods enable\nnew approaches to high-dimensional\, contact-rich manipulation. Next\, I\nwill explore how foundation models can be combined with global\noptimization to reduce modeling burdens and improve generalization of\nmodel-based solvers. I will conclude with an outlook on using\noptimization principles to reshape latent geometry for world-model\nlearning. Together\, these examples illustrate how learning can expand\nthe reach of optimization\, while optimization can bring structure and\nguarantees to data-driven robotics. \nBio:\nFrederike Dümbgen is an assistant professor in the Department of\nMechanical Engineering at Carnegie Mellon University\, where she leads\nthe Trustworthy Robotics\, Intelligence\, and Optimization (TRIO) Lab. Her\nresearch combines optimization and artificial intelligence to develop\nscalable and trustworthy robotic systems\, with a particular emphasis on\nglobal optimization and optimality guarantees.\nBefore joining Carnegie Mellon\, she was a researcher in the WILLOW team\nat Inria Paris and a postdoctoral fellow at the Robotics Institute of\nthe University of Toronto. She holds a Ph.D. in Computer and\nCommunication Sciences from École polytechnique fédérale de Lausanne\n(EPFL)\, Switzerland\, where she also earned her B.Sc. and M.Sc. in\nMechanical Engineering. She was an RSS Pioneer in 2024 and received a\nGoogle Women Techmakers Scholarship in 2020. She serves as a co-chair of\nthe IEEE Robotics and Automation Society’s Technical Committee on\nOptimization for Robotics.
URL:https://www.ri.cmu.edu/event/global-optimization-for-robotics-in-the-age-of-learning/
LOCATION:Newell-Simon Hall 4305
CATEGORIES:Faculty Events
ATTACH;FMTTYPE=image/jpeg:https://www.ri.cmu.edu/app/uploads/2026/10/10-16-26-profile-crop-scaled.jpg
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20261009T143000
DTEND;TZID=America/New_York:20261009T153000
DTSTAMP:20261115T000633
CREATED:20260828T165329Z
LAST-MODIFIED:20261014T133837Z
UID:153343-1791556200-1791559800@www.ri.cmu.edu
SUMMARY:How Can Robots Learn to Interact and Coordinate with Others?
DESCRIPTION:Abstract: Robots in the real world rarely act alone. They share spaces with other robots and with people\, and succeeding there means learning to interact and coordinate with them. This talk looks at what makes interaction hard and how we might get there. I start with multi-agent reinforcement learning\, where challenges like credit assignment make coordination difficult to learn\, and show how the reasoning capabilities of foundation models can help coach robots to learn coordination. Coordination doesn’t have to be learned through trial and error though\, so I then look at learning it from demonstrations. Imitating interacting agents turns out to be quite different from imitating a single one but it can still build on single-agent imitation. I then close by reversing the question: rather than asking what single-agent methods can do for multi-agent problems\, I ask what multi-agent thinking can do for a single robot and show that continual learning\, for example\, can be cast as a consensus optimization problem which results in a natural way to balance learning new tasks against retaining old ones.\n\nBio: Negar Mehr is an Assistant Professor of Mechanical Engineering at UC Berkeley\, where she directs the Berkeley Intelligent Control (ICON) Lab. Her research focuses on multi-agent interactions. Her lab develops learning-based methods grounded in game theory and control to enable robots to interact safely and intelligently with humans and other autonomous agents. She received her Ph.D. in Mechanical Engineering from UC Berkeley in 2019 and was a postdoctoral scholar in Aeronautics and Astronautics at Stanford University. Before returning to Berkeley\, she was an Assistant Professor of Aerospace Engineering at the University of Illinois Urbana-Champaign. She is a recipient of the ONR Young Investigator Award\, the NSF CAREER Award\, the ASME Rising Star Award\, and the IEEE ITSS Best Ph.D. Dissertation Award.
URL:https://www.ri.cmu.edu/event/ri-seminar-with-negar-mehr/
LOCATION:1403 Tepper School Building
CATEGORIES:RI Seminar,Seminar
ATTACH;FMTTYPE=image/jpeg:https://www.ri.cmu.edu/app/uploads/2026/08/negar-headshot-scaled-1.jpg
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20261009T133000
DTEND;TZID=America/New_York:20261009T143000
DTSTAMP:20261115T000633
CREATED:20260924T142654Z
LAST-MODIFIED:20260924T142654Z
UID:153809-1791552600-1791556200@www.ri.cmu.edu
SUMMARY:RI Faculty Business Meeting
DESCRIPTION:Meeting for RI Faculty.\nIn person location – NSH 4305.\nZoom link available via calendar invite.
URL:https://www.ri.cmu.edu/event/ri-faculty-business-meeting-33-2-2-2-2-2-2-2-2-2-2-2-2-2-2-2-2-2-2-2-2/
LOCATION:Newell-Simon Hall 4305
CATEGORIES:Faculty Events
ATTACH;FMTTYPE=image/png:https://www.ri.cmu.edu/app/uploads/2023/11/ri-new-mark-512-512-transparent.png
ORGANIZER;CN="RI Director's Office":MAILTO:lynnetta@cs.cmu.edu
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20261009T113000
DTEND;TZID=America/New_York:20261009T123000
DTSTAMP:20261115T000633
CREATED:20261001T200734Z
LAST-MODIFIED:20261001T200734Z
UID:153873-1791545400-1791549000@www.ri.cmu.edu
SUMMARY:RI PhD Speaking Qual - Ryan Schuerkamp
DESCRIPTION:Date: Friday\, October 9\nTime: 11:30 am – 12:30 pm\nRoom location: GHC 4405\nZoom link: https://cmu.zoom.us/j/9873588831\n\nTitle: AdaSoS: Escaping the Exponential Cost of Higher-Order Sum-of-Squares Relaxations\n\nAbstract:\nMany problems in robotics\, power systems\, and machine learning are nonconvex polynomial optimization problems; local solvers return an answer but cannot tell you how far it is from the best one. The Sum-of-Squares (SoS) hierarchy fixes this with a sequence of convex semidefinite relaxations that give certified lower bounds and\, at high enough degree\, the global optimum. The catch is cost. Each step up the hierarchy grows the relaxation exponentially\, which puts higher-order SoS out of reach for real problems such as AC optimal power flow (AC-OPF)\, the problem of scheduling generators on the electric grid.\n\nThis talk presents Adaptive-SoS (AdaSoS)\, which gets the tightness of a higher-order relaxation without building it. AdaSoS checks whether the current low-order solution could extend to a valid higher-degree one. When it cannot\, the failure itself identifies which polynomial directions are missing\, and AdaSoS adds only those. We prove that AdaSoS reaches the full higher-order bound in finitely many steps. On our largest AC-OPF relaxation\, it certifies the global optimum 13× faster than the full higher-order relaxation\, with a 77% smaller semidefinite block. At the highest degree we solve\, it matches or improves on the bounds of existing approaches such as TSSOS\, CS-TSSOS\, and the adaptive hierarchy of Josz and Molzahn\, typically with a much smaller semidefinite block. AdaSoS can also start from cheaper sparse relaxations\, which lets it scale: on the IEEE 300-bus system\, where the dense higher-order relaxation does not fit in memory\, AdaSoS started from a sparse degree-1 relaxation tightens its bound to within 0.14% of the best known AC-OPF solution.\n\nCommittee members:\nDr. Geoff Gordon (advisor)\nDr. Drew Bagnell\nDr. Changliu Liu\nXinyu Li
URL:https://www.ri.cmu.edu/event/ri-phd-speaking-qual-ryan-schuerkamp/
LOCATION:Gates Hillman Center 4405
CATEGORIES:PhD Speaking Qualifier,Student Talks
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20261007T153000
DTEND;TZID=America/New_York:20261007T170000
DTSTAMP:20261115T000633
CREATED:20260929T175722Z
LAST-MODIFIED:20260929T180200Z
UID:153850-1791387000-1791392400@www.ri.cmu.edu
SUMMARY:RI PhD Thesis Proposal - Shashwat Singh
DESCRIPTION:RI CALENDAR EVENT \n\n\n\n\n\n\n\n\n\n\nWho: Shashwat Singh\nType: RI Thesis Proposal\nDate: Wednesday\, October 7\, 2026\nTime: 03:30PM – 05:00PM (ET)\nLocation: NSH 4305 \nZoom link: Here \nTitle: Extending Robot Mobility Through Mechanical Adaptation \nAbstract: \nUnstructured terrain presents diverse challenges to robot locomotion\, including obstacles\, deformable surfaces\, and confined passages. Terrain conditions also change with time\, as rainfall alters surface traction or shifting debris and rubble block previously traversable terrain. Addressing these challenges requires robots to adapt how they move and interact with their environments. In this thesis\, I investigate how mechanical adaptation can extend the mobility of centimeter-scale robots. \nFirst\, I study how switching between complementary locomotion modes extends mobility. A springtail-inspired microrobot (2.1 cm long; 0.98 g)\, combines crawling and jumping using a single actuator to overcome obstacles. TerraSkipper (5.8 cm long; 28 g) uses impulsive skipping to traverse sand and mud where fin-based crawling is ineffective\, while the RESCUE Jumper (8.2cm long; 125 g) reuses its wheel actuators to jump onto steps it cannot climb by rolling. \nI further examine how physical collaboration extends mobility rather than integrating every capability into a single robot platform. A centimeter-scale RESCUE roller (9 cm long; 100g) couples with other rollers and with a soft growing robot\, also known as Vine robot to share strength and power\, and improve mobility. In this work\, I used RESCUE Rollers to manipulate a Vine robot in a two-dimensional plane\, expanding its operational workspace. To extend this capability even further\, I propose developing Drone Roller (19.4cm long; 200g)\, a compact aerial–ground version of the RESCUE Roller that can fly over obstacles it cannot climb and guide the Vine robot in three-dimensional space. \nFinally\, I study how incorporating an adaptive spine into a robot’s body extends mobility. I designed BaSiL (31 cm long\, 240g)\, a wheeled robot whose tendon-driven beaded spine combines passive compliance with active bending. Actuating the spine in predefined sequences increases the height of steps the robot climbed more than sixfold compared with a rigidly constrained spine. Furthermore\, I propose to investigate continuous morphological adaptation using the All Wheel Morph (AWM) robot (48 cm long\, 4640g)\, whose rotating wheels vary continuously in shape between round wheels and leg-like configurations. I will evaluate AWM through controlled laboratory experiments on steps\, sand\, and mud\, along with real-world experiments. These experiments will test whether the optimal morphology varies continuously with terrain conditions or changes abruptly between distinct wheel or leg configurations. \nWith the proposed work\, I aim to understand how robot morphology can be matched to different terrain conditions\, explore the benefits and trade-offs of adaptation\, and understand design guidelines for improving terrain traversal. \n  \nThesis Committee Members:\nZeynep Temel (Chair)\nSarah Bergbreiter\nAaron Johnson\nPakpong Chirarattananon (University of Toronto) \nDraft of the Thesis Proposal Document
URL:https://www.ri.cmu.edu/event/ri-phd-thesis-proposal-shashwat-singh/
LOCATION:Newell-Simon Hall 4305
CATEGORIES:PhD Thesis Proposal,Student Talks
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20261005T153000
DTEND;TZID=America/New_York:20261005T163000
DTSTAMP:20261115T000633
CREATED:20260917T193616Z
LAST-MODIFIED:20260917T193616Z
UID:153744-1791214200-1791217800@www.ri.cmu.edu
SUMMARY:Video Intelligence for Understanding Behavior and Guiding Action
DESCRIPTION:Abstract:   ​Modern video-language models have become remarkably good at describing what is happening in a video. Yet their ability to understand the behavior they observe\, and to act on that understanding\, remains limited. This talk presents our group’s work on building these missing capabilities. I will begin with foundational methods for scalable video perception\, from space-time attention to hierarchical representations and active search over long videos. I will then introduce strategic video intelligence\, using team sports as a natural microworld in which cooperative and competitive behavior can be studied rigorously\, and present two complementary lines of work within it. The first investigates whether video generation models can simulate strategic multi-agent behavior well enough to support better decisions. The second examines how agents can study behavioral histories across large multimodal corpora\, compiling repeated analytical work into reusable representations\, abstractions\, and procedures. Finally\, I will present WatchAct\, a benchmark for robot manipulation grounded in observed human behavior\, where a robot watches a person and then has to act on what it saw. Together\, these works move video models from perceiving behavior\, to understanding it\, to guiding action. \nBio:  ​Gedas Bertasius is an Assistant Professor in the Department of Computer Science at the University of North Carolina at Chapel Hill. His research focuses on video understanding and multimodal AI\, particularly on systems that can perceive\, reason about\, and act on behavior observed in video. He received his Ph.D. in Computer Science from the University of Pennsylvania and was a postdoctoral researcher at Facebook AI Research before joining UNC. He is the recipient of the NSF CAREER Award\, four Sony Research Awards\, two Meta Research Awards\, and the Egocentric Vision Distinguished Paper Award. He was also a CVPR Best Paper Award finalist. Prior to his research career\, he played professional basketball in Lithuania\, an experience that still shapes his work on understanding strategic behavior in video. \nHomepage: https://www.gedasbertasius.com/\n\nSponsor:\nThe VASC seminar is generously sponsored by HeyGen\, an all-in-one\nAI-powered video generation platform that leverages advances in\ncomputer vision\, generative modeling\, and multimodal learning to make\nhigh-quality video creation both scalable and accessible.
URL:https://www.ri.cmu.edu/event/video-intelligence-for-understanding-behavior-and-guiding-action/
LOCATION:3305 Newell-Simon Hall
CATEGORIES:Seminar,VASC Seminar
ATTACH;FMTTYPE=image/jpeg:https://www.ri.cmu.edu/app/uploads/2026/09/10-5-26-scaled.jpg
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20261002T143000
DTEND;TZID=America/New_York:20261002T153000
DTSTAMP:20261115T000633
CREATED:20260828T165026Z
LAST-MODIFIED:20260929T180726Z
UID:153339-1790951400-1790955000@www.ri.cmu.edu
SUMMARY:Shaping the Body of Robots with Soft Skin and Muscles
DESCRIPTION:Abstract:\nJust as skin and muscle give living creatures their ability to sense\, move\, and adapt\, soft sensors and actuators are giving shape to a new kind of robot body. Built from flexible\, compliant materials\, soft robots can closely mimic the movements of biological systems\, a significant step beyond conventional rigid machines. This talk explores recent technologies for designing and fabricating the two essential components of that body: artificial skin that lets robots feel their surroundings\, and artificial muscles that let them move with natural\, fluid motion. We will discuss how these components are engineered for responsiveness and reliability\, how their signals can be interpreted despite the complex\, nonlinear behavior of soft structures\, and how sensing and actuation can be co-designed into a single integrated system. By integrating skin and muscle into a unified soft body\, we can develop a new generation of robots capable of seamless and safe interaction in human environments.\n\nBio:\nYong-Lae Park (Member\, IEEE) received the Ph.D. degree in mechanical engineering from Stanford University\, Stanford\, CA\, USA\, in 2010.\nHe is currently a Professor with the Department of Mechanical Engineering\, Seoul National University\, Seoul\, South Korea. Prior to joining SNU\, he was an Assistant Professor in the Robotics Institute\, Carnegie Mellon University\, Pittsburgh\, PA\, USA (2013–2017) and a Technology Development Fellow in the Wyss Institute for Biologically Inspired Engineering\, Harvard University\, Cambridge\, MA\, USA (2010–2013). His research interests include artificial skin sensors and muscle actuators\, soft robots\, wearable robots\, rehabilitation robots\, and medical robots.\nProf. Park is the recipient of  the Best Conference Paper Award from the IEEE International Conference of Soft Robotics (2019)\, the Best Application Paper Award from the IEEE Transactions on Haptics (2019)\, and the Best Paper Award from the IEEE Sensors Journal (2013). He is currently serving as a Senior Editor in IEEE Robotics and Automation Letters\, an Associate Editor in Soft Robotics and npj Robotics.
URL:https://www.ri.cmu.edu/event/shaping-the-body-of-robots-with-soft-skin-and-muscles/
LOCATION:1403 Tepper School Building
CATEGORIES:RI Seminar,Seminar
ATTACH;FMTTYPE=image/png:https://www.ri.cmu.edu/app/uploads/2026/08/ylpark_headshot2.png
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20260930T160000
DTEND;TZID=America/New_York:20260930T170000
DTSTAMP:20261115T000633
CREATED:20261001T142902Z
LAST-MODIFIED:20261001T142902Z
UID:153867-1790784000-1790787600@www.ri.cmu.edu
SUMMARY:RI PhD Speaking Qual - Jianjin Xu
DESCRIPTION:Date: Wednesday\, Sep 30\, 2026Time: 16:00 – 17:00 PMLocation: NSH 4305Zoom: https://cmu.zoom.us/j/97595860992?pwd=voNXd0zhn8HdqFakFOF3At8y0NbydV.1Title: Efficient 3D Avatar Reconstruction with Geometric Guidance \nAbstract:3D animatable human avatars are widely used in film making\, game characters\, and telepresence. To create these avatars efficiently\, researchers propose to reconstruct the avatar from several input images with a feedforward network. These networks are mostly transformers trained with large proprietary data and thousands of H100 GPU hours. However\, do we really need such a scale of data and compute for this task? \nOur answer is no. In this talk\, we present ARG-Avatar\, a lightweight network with only 68M trainable parameters\, yet achieves SOTA performance on OOD testing data with 13x less training compute to the best baseline. We will go through the two core components of ARG-Avatar. The first is FACRoPE\, which inject geometric guidance into the attention with RoPE mechanism. We propose a novel coordinate formulation named Foreground Avatar Coordinates (FAC)\, to associate let the network tokens focus on its corresponding image regions. The second is Intermediate Token Rendering (ITR)\, which decodes a coarse avatar from intermediate network tokens during forward pass. We show that these components are all beneficial in the ablation study. In conclusion\, we show that by properly injecting geometric into attention\, we make an architecture that learns more efficiently and effectively for feedforward avatar reconstruction. \n\n—\n \n\nBest regards\, \nJianjin Xu.\nCarnegie Mellon University\, Ph.D. in Robotics\nhttps://atlantixjj.github.io/
URL:https://www.ri.cmu.edu/event/ri-phd-speaking-qual-jianjin-xu/
LOCATION:Newell-Simon Hall 4305
CATEGORIES:PhD Speaking Qualifier,Student Talks
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20260928T153000
DTEND;TZID=America/New_York:20260928T163000
DTSTAMP:20261115T000633
CREATED:20260914T232638Z
LAST-MODIFIED:20260914T232638Z
UID:153556-1790609400-1790613000@www.ri.cmu.edu
SUMMARY:From Capturing People to Teaching Robots
DESCRIPTION:Abstract:   Equipping AI and robotic systems with the ability to understand human behavior is essential for enabling them to assist people across a wide range of everyday applications. This need is more pressing than ever: the heaviest consumers of such knowledge are no longer perception systems alone\, but robot policies that must learn to act in the physical world. Yet the high-quality 3D human motion data required to learn this knowledge remains extremely scarce. \nIn this talk\, I will present our lab’s efforts to scale and enrich 3D human motion data by capturing everyday movements and natural human-object interactions\, with the ultimate goal of teaching robots to move like humans. \nI will first introduce our multi-year effort in building multi-camera capture systems\, from Panoptic Studio to ParaHome\, and most recently OmniRoboHome\, a new system designed to capture human-object interactions in natural home environments. Next\, I will present a complementary direction: learning everyday interactions and affordances from generative image and video models\, which offer a scalable source of human behavior priors that capture systems alone cannot reach. Finally\, I will discuss the missing pieces that vision and image models cannot provide\, including physics\, contact\, and the gap across diverse embodiments\, together with our recent efforts to fill them. \nBio:  Hanbyul Joo is an associate professor at Seoul National University (SNU) in the Department of Computer Science and Engineering. Before joining SNU\, Hanbyul was a Research Scientist at Facebook AI Research (FAIR)\, Menlo Park. Hanbyul received his PhD from the Robotics Institute at Carnegie Mellon University. He is a recipient of the Samsung Scholarship\, the Okawa Foundation Research Grant\, and the Best Student Paper Award at CVPR 2018. \nHomepage: https://jhugestar.github.io/ \nSponsor:\nThe VASC seminar is generously sponsored by HeyGen\, an all-in-one\nAI-powered video generation platform that leverages advances in\ncomputer vision\, generative modeling\, and multimodal learning to make\nhigh-quality video creation both scalable and accessible.
URL:https://www.ri.cmu.edu/event/from-capturing-people-to-teaching-robots/
LOCATION:Newell-Simon Hall 3305
CATEGORIES:Seminar,VASC Seminar
ATTACH;FMTTYPE=image/jpeg:https://www.ri.cmu.edu/app/uploads/2026/09/9-28-26-han_dec_2017-4.jpg
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20260925T110000
DTEND;TZID=America/New_York:20260925T130000
DTSTAMP:20261115T000633
CREATED:20260911T143145Z
LAST-MODIFIED:20260911T164700Z
UID:153524-1790334000-1790341200@www.ri.cmu.edu
SUMMARY:Adaptive Cameras: Bridging Novel Sensors and Robot Perception
DESCRIPTION:Abstract:  Most cameras on robots today capture images without considering scene content. In contrast\, animal eyes have fast mechanical movements that control how the scene is imaged in detail by the fovea\, where visual acuity is highest. The prevalence of active vision during biological imaging\, and the wide variety of it\, makes it very clear that this is an effective visual design strategy for robot vision. In this talk\, I will cover our recent work on creating *both* new camera designs and novel robot perception algorithms to enable adaptive and selective active vision and imaging inside cameras and sensors. \nBio:  Sanjeev J. Koppal is an Associate Professor at the University of Florida’s Electrical and Computer Engineering Department and is a Kent and Linda Fuchs Faculty Fellow. He also held a UF Term Professorship for 2021-23. Sanjeev is the Director of the FOCUS Lab at UF. Since 2022\, Sanjeev has been an Amazon Scholar with Amazon Robotics. Prior to joining UF\, he was a researcher at the Texas Instruments Imaging R&D lab. Sanjeev obtained his Masters and Ph.D. degrees from the Robotics Institute at Carnegie Mellon University. After CMU\, he was a postdoctoral research associate in the School of Engineering and Applied Sciences at Harvard University. He received his B.S. degree from the University of Southern California in 2003 as a Trustee Scholar. He is a co-author on best student paper awards for ECCV 2016 and NEMS 2018\, and work from his FOCUS lab was a CVPR 2019 best-paper finalist. Sanjeev won an NSF CAREER award in 2020 and is an IEEE Senior Member and an Optica Senior Member. He won a UF ECE Department Teaching Award in 2024. His interests span computer vision\, computational photography and optics\, novel cameras and sensors\, 3D reconstruction\, physics-based vision\, and active illumination.\n\nHomepage: https://focus.ece.ufl.edu/ \nSponsor:\nThe VASC seminar is generously sponsored by HeyGen\, an all-in-one\nAI-powered video generation platform that leverages advances in\ncomputer vision\, generative modeling\, and multimodal learning to make\nhigh-quality video creation both scalable and accessible.
URL:https://www.ri.cmu.edu/event/adaptive-cameras-bridging-novel-sensors-and-robot-perception/
LOCATION:Newell-Simon Hall 4305
CATEGORIES:Seminar,VASC Seminar
ATTACH;FMTTYPE=image/jpeg:https://www.ri.cmu.edu/app/uploads/2026/09/9-25-26-SanjeevKoppal.jpg
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20260923T141500
DTEND;TZID=America/New_York:20260923T163000
DTSTAMP:20261115T000633
CREATED:20260921T143237Z
LAST-MODIFIED:20260923T173509Z
UID:153762-1790172900-1790181000@www.ri.cmu.edu
SUMMARY:RI PhD Thesis Defense - Bardienus Duisterhof
DESCRIPTION:Date: 9/23/2026 \nTime: 2:00 PM – 4:00 PM \nZoom: https://cmu.zoom.us/j/98474543738?pwd=PVQo6JfRsWvuB9eWLqbatVivVv7CTj.1 \nLocation: NSH 4305 \nType: PhD Thesis Defense \nWho: Bardienus Pieter Duisterhof \nTitle: Spatiotemporal World Models for Robot Manipulation \nAbstract: \nRobot manipulation aims to automate tasks that are too dull\, dirty\, or dangerous for humans. This future requires remarkable resource efficiency: robots must adapt to new tasks with limited data and compute while meeting stringent performance requirements. Current systems can succeed at dexterous tasks but require substantial resources to meet industrial standards. One possible explanation for this inefficiency is that frontier models learn directly from RGB images\, which are typically dominated by content irrelevant to robots. Previous work has addressed this problem through task-specific feature engineering\, improving efficiency by focusing on task-relevant information. Can we construct similarly focused representations that remain scalable and broadly applicable? This thesis explores spatiotemporal representations—representations of 3D geometry and motion over time—focusing on their reconstruction\, generation\, and robot applications. \nThe first part of this thesis addresses unconstrained spatiotemporal reconstruction. Robots may benefit from representations that place past observations in a precise spatial and temporal context. We contribute methods that improve 3D reconstruction and calibration from arbitrary image sets and lens configurations. We show that calibrated multi-camera setups and neural rendering yield precise reconstruction in dynamic scenes\, including highly deformable objects such as cloth. \nIn the second part of this thesis\, we investigate learning spatiotemporal generative priors for robot manipulation. Humans can infer plausible geometry and dynamics from a single observation. Can we instill similar priors into robots? We contribute methods that can infer depth maps\, complete object geometry\, and predict object dynamics. With Modality Forcing\, we investigate text-to-image pre-training as a way to learn geometric priors. With PointZero\, we use 3D point track completion to learn spatiotemporal priors without any robot data. \nThe final part of this thesis considers spatiotemporal world models applied to robot manipulation. First\, we show that PointZero improves performance in robot manipulation tasks\, including imitation learning and action-conditioned dynamics prediction. Next\, we investigate how predicting future scene states can guide action generation in world-action models (WAMs). In 3PoinTr\, we show that 3D point tracks can serve as compact task plans that support transfer from human demonstrations to robot execution. In ModAR\, we autoregressively denoise multiple future modalities and robot actions\, achieving the best performance among the WAM formulations tested. We systematically study which modalities contribute most to manipulation performance and find that predicting RGB images provides no consistent additional benefit at the scale studied. \nTogether\, these works connect reconstruction\, learned geometric and dynamics priors\, and action generation to study how spatiotemporal representations can support efficient\, scalable robot learning. \nCommittee:  \nJeffrey Ichnowski (Chair)\nDeva Ramanan\nShubham Tulsiani\nAbhishek Gupta (University of Washington)
URL:https://www.ri.cmu.edu/event/ri-phd-thesis-defense-bardienus-duisterhof-2/
LOCATION:Newell-Simon Hall 4305
CATEGORIES:PhD Thesis Defense,Student Talks
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20260921T153000
DTEND;TZID=America/New_York:20260921T163000
DTSTAMP:20261115T000633
CREATED:20260908T153454Z
LAST-MODIFIED:20260908T153454Z
UID:153488-1790004600-1790008200@www.ri.cmu.edu
SUMMARY:Decision-Making in a World of Latent Particles
DESCRIPTION:Abstract: Robots must often make decisions in scenes containing many objects: they need to identify what is present\, understand where objects are\, predict how they will interact\, and choose actions accordingly. Learning these capabilities directly from pixels is challenging\, especially when the number and arrangement of objects can change from one scene to another. \nIn this talk\, I will presentDeep Latent Particles (DLP)\, aself-supervisedobject-centric representation that describes a visual scene as a set of compact latent particles. Each particle captures the location and visual properties of a discovered object or object part\, providing an interpretable bridge between raw images and multi-object decision-making. \nI will show how DLP can serve as a representation for learning robotic policies from online reinforcement learning\, offline data\, and demonstrations. In particular\, policies built on these representations can generalize compositionally to scenes containing more objects than were present during training. \nI will then introduce Latent Particle World Models(ICLR 2026 Oral)\, which learn to predict how collections of latent particles evolve over time. These object-centric world models support multi-view observations and flexible conditioning\, enabling prediction and decision-making in rich visual environments. I will discuss how this perspective connects to diffusion-based policies and world action models (WAMs)\, and conclude with a look toward self-supervised 3D object-centric learning for robots that can perceive\, predict\, and act in three-dimensional worlds. and which will be quietly subsumed by the next scale-up. \n  \nBio: Tal Daniel is a Postdoctoral Fellow at Carnegie Mellon University’s Robotics Institute\, working with Prof. Deepak Pathak and Prof. David Held. He received his Ph.D. in Electrical and Computer Engineering from the Technion\, advised by Prof. Aviv Tamar. His research spans self-supervised and object-centric representation learning\, generative modeling\, reinforcement learning\, and robotics\, with a focus on learning representations and world models. \n  \nHomepage: https://taldatech.github.io \n  \nSponsor: \nThe VASC seminar is generously sponsored by HeyGen\, an all-in-oneAI-powered video generation platform that leverages advances incomputer vision\, generative modeling\, and multimodal learning to makehigh-quality video creation both scalable and accessible.
URL:https://www.ri.cmu.edu/event/decision-making-in-a-world-of-latent-particles/
LOCATION:3305 Newell-Simon Hall
CATEGORIES:Seminar,VASC Seminar
ATTACH;FMTTYPE=image/png:https://www.ri.cmu.edu/app/uploads/2026/09/9-21-26-tal_daniel.png
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20260918T150000
DTEND;TZID=America/New_York:20260918T163000
DTSTAMP:20261115T000633
CREATED:20260914T133740Z
LAST-MODIFIED:20260914T133740Z
UID:153537-1789743600-1789749000@www.ri.cmu.edu
SUMMARY:A Reactive Vibration Compensation System for Discrete Material Deposition Processes
DESCRIPTION:Abstract:  \n\n\n\n\nUltra Large Format Deposition (ULF-D) systems perform precision material deposition over large workspaces\, yet their extended and mechanically compliant structures can produce configuration-dependent vibration at the tool mounted to the robot’s end-effector. In inkjet printing\, vibration perturbs the deposition tool such that each discrete deposit (a single ink drop) lands displaced from its intended location on the target surface\, thus creating errors* in the deposited pattern. Conventional motion-control vibration compensation focuses on suppressing the vibration and may be limited by uncertainty in vibration dynamics models\, insufficient control bandwidth\, or restricted access to the underlying motion controller. \n\n\nAn alternative approach to reducing deposition error is process control\, which monitors and adjusts process variables to achieve a desired output. In inkjet printing\, each drop is released by a digital trigger signal\, so the instant the trigger fires determines where along the path the drop lands. This timing can be controlled independently of the robot’s trajectory. \n\n\nRather than applying corrective robot motion\, this thesis presents a process-control framework that adapts deposition-event timing to compensate for errors caused by bounded tool vibration. The framework combines a synchronized\, low-latency sensing and actuation architecture; a tool state estimation pipeline that fuses high-rate (1 kHz) inertial measurements with lower-rate (100 Hz) global pose measurements; and a reactive scheduler that triggers deposition events according to the estimated progression of the tool along its planned path. Together\, these components decrease deposition error by adjusting deposition timing with high-rate vibrating tool state feedback without modifying the robot’s motion-control stack. \n\n\nExperiments on a ULF-D emulation testbench evaluate the framework across multiple trajectories under both structured and unstructured vibration conditions. The proposed scheduler reduces mean deposition error by 67% to 91% relative to fixed-time firing\, keeping error between 0.1 mm and 1.4 mm across all tested conditions despite tool vibration amplitudes of up to 11 mm.\n\n\n\n*The mean absolute difference between the desired and measured spacing of adjacent line features in a deposition pattern. \n\n\n\n\n\nCommittee:  \n\n\nDr. Howie Choset (advisor) \n\n\nDr. Wennie Tabib \n\n\nDarwin Mick
URL:https://www.ri.cmu.edu/event/a-reactive-vibration-compensation-system-for-discrete-material-deposition-processes/
LOCATION:Gates Hillman Center 4405
CATEGORIES:MSR Thesis Presentation,Student Talks
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20260918T113000
DTEND;TZID=America/New_York:20260918T123000
DTSTAMP:20261115T000633
CREATED:20260915T180113Z
LAST-MODIFIED:20260915T201300Z
UID:153571-1789731000-1789734600@www.ri.cmu.edu
SUMMARY:First Light: Spaceborne multi-view computational tomography (CT)
DESCRIPTION:Abstract:  We devise multiview tomography from space. This talk will show the First Light of CloudCT\, the very first image taken since we launched our first satellite. This opens opportunities for scientific observations and technologies\,  can transform climate research and potentially even affect medical X-ray CT.  Tasks and solutions involve both novel imaging hardware and computational algorithms\, based on machine learning and differential rendering. The key idea is that advanced computing enable computed tomography of volumetric scenes\, based scattered radiation.  CloudCT\, funded by the ERC\, involves 10 nano-satellites to fly in an unprecedented formation\, to capture the same scene (cloud fields) from multiple views simultaneously. We encounter challenges of  polarimetric self-calibration in orbit\, and estimation of 3D volumetric distribution of microphysical properties. \nBio:  Prof. Yoav Schechner\, heads the Asher Space Research Inst. and is a faculty member at the Viterbi Faculty of Electrical and Computer  Engineering\, Technion. He is a Principal Investigator and Coordinator of the ERC CloudCT space project\, and is on the science team of two additional space-imaging projects\, C3IEL (by ISA/CNES) and MAIA (by NASA/ASI). He was a Research Scientist at Columbia U.\, and a Visiting Associate in Caltech\, the Jet Propulsion Laboratory (JPL) and MIT.  His main interests are diverse forms of computational imaging.  We won the Best Paper Award\, ICCP 2013+2018\, Best Student Paper Award\, CVPR 2017\, Distinguished Teaching Awards\, Technion\, Fumio Okano Best 3D Paper Award\, and others. He trained in physics (BA+MSc) and EE (PhD). He is the Mark and Diane Seiden Chair in Science and a Landau Fellow – supported by the Taub Foundation. \nSponsor:\nThe VASC seminar is generously sponsored by HeyGen\, an all-in-one\nAI-powered video generation platform that leverages advances in\ncomputer vision\, generative modeling\, and multimodal learning to make\nhigh-quality video creation both scalable and accessible.
URL:https://www.ri.cmu.edu/event/first-light-spaceborne-multi-view-computational-tomography-ct/
LOCATION:Newell-Simon Hall 4305
CATEGORIES:Seminar,VASC Seminar
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20260918T100000
DTEND;TZID=America/New_York:20260918T113000
DTSTAMP:20261115T000633
CREATED:20260908T155758Z
LAST-MODIFIED:20260908T155947Z
UID:153494-1789725600-1789731000@www.ri.cmu.edu
SUMMARY:RI PhD Thesis Defense - Ingrid Navarro Anaya
DESCRIPTION:Date: September 18th\, 2026 \nTime: 10:00 AM (ET) \nZoom: link \nLocation: NSH 4305 \nType: PhD Thesis Defense \nWho: Ingrid Navarro Anaya \n  \nTitle: Towards Generalizable Motion Prediction under Distribution Shifts \n  \nAbstract: \nAutonomous robots are increasingly expected to operate in dynamic\, human-centered environments. To do so safely and efficiently\, they must reason about how people move and interact. In domains like driving\, social navigation\, and aviation\, a common approach is to learn models of human motion directly from recorded data and use the resulting priors to inform downstream systems like simulators and planning stacks. \n  \nDespite the growing availability of datasets\, benchmarks\, and modeling techniques\, state-of-the-art methods remain unreliable for real-world deployment\, often generalizing poorly to novel environments and rare events. Much of this stems from recorded datasets covering few environments and few safety-relevant events relative to what a deployed system will ultimately encounter. Broadly\, the field has addressed this in four main ways: validating autonomy stacks on the road\, collecting more data\, synthesizing relevant scenarios\, and adapting at test time. These are all valuable and necessary strategies\, but each is bounded by risk\, cost\, the sim-to-real gap\, or the difficulty of reliably detecting and adapting to a shift\, respectively. \n  \nThese strategies share the premise that the data we hold is insufficient. This dissertation argues that such data is also underexploited and thus pursues a complementary direction\, asking how much actionable signal existing datasets already contain but current practice overlooks. We do so through a recurring paradigm we call scenario characterization: describing a scenario in terms of a property of interest and acting on that description downstream. We use this paradigm in three ways. The first is for guidance and abstraction\, shaping what a model trains or optimizes. The second is for mining and benchmarking\, determining what a model trains on and what is withheld to test it. The third is for analysis\, fixing the basis on which results are studied. We apply these uses across two settings: in-distribution generalization\, which draws mainly on guidance and abstraction\, and generalization under distribution shift\, which draws on all three and is where the main contributions concentrate. \n  \nWe further argue that evidence for generalization is typically gathered only within a single domain\, so a claim that holds there is rarely challenged elsewhere. This is largely because\, outside autonomous driving\, no motion domain offers comparable infrastructure and scale. This dissertation therefore contributes aviation as a testbed\, introducing a large-scale framework and dataset for airport surface movement forecasting. \n  \nThrough this framework and our experimental settings\, we expose generalization failures that would otherwise have remained hidden in aggregate metrics. We also enable reading aviation and driving on a common basis\, hinting at what transfers and what does not. Finally\, we also provide early evidence that the signal recovered through our framework may help systems gain the risk awareness needed to handle real-world critical events. \n  \nThesis Committee: \nJean Oh (co-chair) \nJonathan Francis (co-chair) \nSebastian Scherer \nAndrea Bajcsy \nAlexandre Alahi (EPFL)
URL:https://www.ri.cmu.edu/event/ri-phd-thesis-defense-ingrid-navarro-anaya/
LOCATION:Newell-Simon Hall 4305
CATEGORIES:PhD Thesis Defense,Student Talks
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20260915T153000
DTEND;TZID=America/New_York:20260915T170000
DTSTAMP:20261115T000633
CREATED:20260903T124903Z
LAST-MODIFIED:20260916T130809Z
UID:153443-1789486200-1789491600@www.ri.cmu.edu
SUMMARY:Making is Decision Making
DESCRIPTION:Abstract:  Human creation of high-quality content requires making decisions – from coarse\, high-level decisions about content and style\, to precise low-level decisions about the color of an individual pixel. Modern generative AI promises to serve a collaborative assistant capable of executing design decisions specified in simple text prompts into high-quality content. Yet today’s AI systems are poor collaborators. They frequently misinterpret user intent\, while users lack a predictive conceptual model of how an AI will interpret a prompt or why it produces a particular result. Without shared conceptual grounding\, collaboration devolves into trial and error\, with users repeatedly rewriting prompts\, using the AI to generate a result and then adjusting the prompt to try again\, in the hope of obtaining the desired outcome. In this talk I’ll argue that for generative AI to fulfill its promise we must develop techniques and interfaces that enable users and AI models to establish shared conceptual grounding. I’ll outline two complementary research challenges; First\, we must identify the concepts that human creators commonly use when reasoning about and communicating within a content creation domain.  Here\, I’ll show how we might extend methods from cognitive psychology to elicit\, represent\, and analyze domain-specific conceptual structures.  Second\, we must develop interfaces for teaching these concepts to AI models. While machine learning is rapidly advancing methods for teaching AI new concepts\, I’ll show how adapting these techniques to the diverse domains of human content creation requires new interaction techniques that support the way people think\, communicate\, and create. Finally\, I’ll demonstrate a few implementations of these ideas that we have developed in our group at Stanford. \nBio: Maneesh Agrawala is the Forest Baskett Professor of Computer Science and Director of the Brown Institute for Media Innovation at Stanford University. He is also a consulting AI Scientist at Roblox. He works on computer graphics\, human computer interaction and visualization. His focus is on investigating how cognitive design principles can be used to improve the effectiveness of audio/visual media. The goals of this work are to discover the design principles and then instantiate them in both interactive and automated design tools. Honors include an Okawa Foundation Research Grant (2006)\, an Alfred P. Sloan Foundation Fellowship (2007)\, an NSF CAREER Award (2007)\, a SIGGRAPH Significant New Researcher Award (2008)\, a MacArthur Foundation Fellowship (2009)\, an Allen Distinguished Investigator Award (2014) and induction into the SIGCHI Academy (2021). He was named an ACM Fellow in 2022.
URL:https://www.ri.cmu.edu/event/making-is-decision-making/
LOCATION:Gates-Hillman Center 4401
CATEGORIES:Special Events
ATTACH;FMTTYPE=image/jpeg:https://www.ri.cmu.edu/app/uploads/2026/09/9-15-26-agrawala-macarthur3-head-square.jpg
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20260911T143000
DTEND;TZID=America/New_York:20260911T153000
DTSTAMP:20261115T000633
CREATED:20260828T164708Z
LAST-MODIFIED:20260911T212228Z
UID:153335-1789137000-1789140600@www.ri.cmu.edu
SUMMARY:Augmenting Bee Colonies with Robotics and AI Technologies for Ecosystem Support
DESCRIPTION:Abstract: Earth’s ecosystems are facing a rapid decline in biodiversity\, with honeybees —keystone pollinators critical to ecosystem stability— being among the most affected. The EU-funded RoboRoyale project addresses this crisis by integrating advanced robotics and AI to augment the beehive\, enabling observation at unprecedented resolutions and scales. Featured on the cover of Science Robotics and receiving the 6th Edge of Government Award at the World Government Summit in 2024 our system tracks the Queen’s behaviors\, colony efficiency\, comb states\, and long-term foraging activities\, while advancing micro-robotic intervention capabilities to support hive health. In this talk\, I will discuss the challenges of developing this system\, share key findings regarding complex social interactions\, and explore the future potential of bio-hybrid research. \nBio: Erol Şahin is a Professor of Computer Engineering at Middle East Technical University (METU) and the founding Director of the Center for Robotics and AI (ROMER). Established with over 5 million Euros in funding\, ROMER spans 25\,000 square feet of state-of-the-art facilities\, including prototyping workshops\, specialized research arenas\, and advanced robotic platforms. Dr. Şahin earned his PhD in Cognitive and Neural Systems from Boston University\, following a BSc in Electrical and Electronics Engineering from Bilkent University and an MSc in Computer Engineering from METU. Before assuming his current position\, he worked as postdoctoral researcher at the Université Libre de Bruxelles.   Between 2013 and 2015\, Dr. Sahin spent two years at the Robotics Institute of Carnegie Mellon University during his sabbatical. His research interests include swarm robotics\, robotic learning\, and human-robot interaction—work that has secured more than 2.5 million Euros from the European Union\, TUBITAK\, and industrial partners. Notably\, his contributions to robotic learning were awarded a 53-DOF iCub humanoid platform through the RobotCub project in 2007. Dr. Şahin has edited several journal special issues and books\, currently serves as an Associate Editor for Adaptive Behavior\, and is a member of the Editorial Board for the Swarm Intelligence journal.
URL:https://www.ri.cmu.edu/event/augmenting-bee-colonies-with-robotics-and-ai-technologies-for-ecosystem-support/
LOCATION:1403 Tepper School Building
CATEGORIES:RI Seminar,Seminar
ATTACH;FMTTYPE=image/jpeg:https://www.ri.cmu.edu/app/uploads/2026/08/erol-sahin.jpg
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20260911T120000
DTEND;TZID=America/New_York:20260911T130000
DTSTAMP:20261115T000633
CREATED:20260901T200426Z
LAST-MODIFIED:20260901T200426Z
UID:153397-1789128000-1789131600@www.ri.cmu.edu
SUMMARY:Computational Lensing
DESCRIPTION:Abstract: From the cameras in our phones to the lenses in head-mounted displays\, optics shape both how we capture the world and how we experience virtual reality. Most conventional lenses are designed to bring a single plane into focus. In this talk\, we will discuss a new class of computational lens—referred to as a Split-Lohmann lens—that provides spatially varying control over focal length. This capability is achieved by combining a phase-only spatial light modulator with the cubic phase plates used in Lohmann/Alvarez focus-tunable lenses. The resulting computational lens enables new imaging and display capabilities\, including the ability to (i) make a flat display appear to have three-dimensional shape\, and (ii) capture all-in-focus images of highly non-planar scenes.
URL:https://www.ri.cmu.edu/event/computational-lensing/
LOCATION:Newell-Simon Hall 4305
CATEGORIES:Faculty Events
END:VEVENT
END:VCALENDAR