![Center for Intelligent Systems CIS EPFL](/img/default-banner.jpg)
- Видео 157
- Просмотров 131 546
Center for Intelligent Systems CIS EPFL
Швейцария
Добавлен 1 окт 2019
Center for Intelligent Systems (CIS)
The EPFL Center for Intelligent Systems brings together researchers working on various aspects of creating intelligent systems.
The Center for Intelligent Systems (CIS) at EPFL, is a joint initiative of the School of Architecture Civil and Environmental
Engineering (ENAC), the School of Computer and Communication Sciences (IC), the School of Basic Sciences (SB), the Schools of Engineering (STI) and the School of Life Sciences (SV) that seeks to advance research and practice in the strategic field of intelligent systems. Such systems are embedded with elements of artificial intelligence and are emerging in response to the convergence of the virtual and physical worlds through the development of autonomous systems, wearable technologies, robotic co-workers, smart devices, drone technology, augmented reality, intelligent houses, digital twins, intelligent manufacturing, and many other applications.
The EPFL Center for Intelligent Systems brings together researchers working on various aspects of creating intelligent systems.
The Center for Intelligent Systems (CIS) at EPFL, is a joint initiative of the School of Architecture Civil and Environmental
Engineering (ENAC), the School of Computer and Communication Sciences (IC), the School of Basic Sciences (SB), the Schools of Engineering (STI) and the School of Life Sciences (SV) that seeks to advance research and practice in the strategic field of intelligent systems. Such systems are embedded with elements of artificial intelligence and are emerging in response to the convergence of the virtual and physical worlds through the development of autonomous systems, wearable technologies, robotic co-workers, smart devices, drone technology, augmented reality, intelligent houses, digital twins, intelligent manufacturing, and many other applications.
Aude Billard (EPFL-CIS):"Linearization and Identification of Multiple-Attractor Dynamical Systems.."
Title: “Linearization and Identification of Multiple-Attractor Dynamical Systems through Laplacian Eigenmaps”
www.epfl.ch/research/domains/cis/epfl-cis-riken-aip-joint-workshop-2023/
www.epfl.ch/research/domains/cis/epfl-cis-riken-aip-joint-workshop-2023/
Просмотров: 156
Видео
Jingfeng Zhang (RIKEN-AIP): “Applications of Adversarial robustness”
Просмотров 8710 месяцев назад
Abstract: When we deploy models trained by standard training (ST), they work well on natural test data. However, those models cannot handle adversarial test data (also known as adversarial examples) that are algorithmically generated by adversarial attacks. An adversarial attack is an algorithm which applies specially designed tiny perturbations on natural data to transform them into adversaria...
M. Ha Quang (RIKEN-AIP)“Information geometry and optimal transport framework for Gaussian processes”
Просмотров 10810 месяцев назад
Abstract: Information geometry (IG) and Optimal transport (OT) have been attracting much research attention in various fields, in particular machine learning and statistics. In this talk, we present results on the generalization of IG and OT distances for finite-dimensional Gaussian measures to the setting of infinite-dimensional Gaussian measures and Gaussian processes. Our focus is on the Ent...
Lenaic Chizat (EPFL-CIS): “A Tractable Barycenter for Probability Measures in Machine Learning”
Просмотров 6610 месяцев назад
Abstract: We introduce a formulation of entropic Wasserstein barycenters that enjoys favorable optimization, approximation, and statistical properties. This barycenter is defined as the unique probability measure that minimizes the sum of entropic optimal transport costs with respect to a family of given probability measures, plus an entropy term. We show that (i) this notion of barycenter is d...
Masatoshi Hamanaka (RIKEN-AIP): “Music Structure Analysis based on Transformers”
Просмотров 3910 месяцев назад
Abstract: We present a time-span tree leveled by the length of the time span. Using the time-span tree of the Generative Theory of Tonal Music, it is possible to reduce notes in a melody, but it is difficult to automate because the priority order of the branches to be reduced is not defined. A similar problem arises in the automation of time-span analysis and melodic morphing. Therefore, we pro...
Martin Jaggi (EPFL-CIS): “Building Blocks for Collaborative and Decentralized Machine Learning”
Просмотров 11810 месяцев назад
www.epfl.ch/research/domains/cis/epfl-cis-riken-aip-joint-workshop-2023/
Masashi Sugiyama (RIKEN-AIP): “Noise Robust Classification”
Просмотров 15110 месяцев назад
Abstract: Supervised learning from noisy output is one of the classic problems in machine learning. While this task is relatively straightforward in regression since independent additive noise cancels out with big data, classification from noisy labels is still a challenging research topic. Recently, it has been shown that when the noise transition matrix which specifies the label flipping prob...
Maryam Kamgarpour (EPFL-CIS): “Learning equilibria in multi-agent systems”
Просмотров 13710 месяцев назад
Abstract: A rising challenge in control of large-scale control systems such as the electricity and the transportation networks is to address autonomous decision making of interacting agents, i.e. the subsystems, with local objectives while ensuring global system safety and performance. In this setting, a Nash equilibrium is a stable solution outcome in the sense that no agent finds it profitabl...
Naonori Ueda (RIKEN-AIP): "Physics-Informed Deep Learning Approach for Modeling Crustal Deformation”
Просмотров 7910 месяцев назад
Abstract: The movement and deformation of the Earth’s crust and upper mantle provide critical insights into the evolution of earthquake processes and future earthquake potentials. Crustal deformation can be modeled by dislocation models that represent earthquake faults in the crust as defects in a continuum medium. In this talk, let me introduce a novel physics-informed deep learning approach t...
Volkan Cevher (EPFL-CIS): “Optimization Challenges in Robust Machine Learning”
Просмотров 18010 месяцев назад
Abstract: “Thanks to neural networks (NNs), faster computation, and massive datasets, machine learning (ML) is under increasing pressure to provide automated solutions to even harder real-world tasks beyond human performance with ever faster response times due to potentially huge technological and societal benefits. Unsurprisingly, the NN learning formulations present a fundamental challenge to...
CIS Network Seminar: #IDIAP - Prof. Lonneke van der Plas
Просмотров 8610 месяцев назад
"Towards creative systems with linguistic modelling" Abstract: Language is predominantly viewed as a means for information exchange in the field of AI and natural language processing (NLP). However, this overlooks the important role language plays as a vehicle for creative thought. This aspect of cognition is unique to mankind, crucial for innovation, and currently under-explored in AI. The fie...
AMLD 2023 AI workshop, "A Super Simple Introduction to AI and Generative AI", Prof. P. Vandergheynst
Просмотров 14410 месяцев назад
« A Super Simple Introduction to AI and Generative AI », by Prof. Pierre Vandergheynst Academic Director of the EPFL Center for Intelligent Systems and Head of the EPFL Signal Processing Laboratory LTS2. On Tuesday, August 29th, 2023, to further the discussions from the main AMLD GenAI Conference 2023, the EPFL has brought together another group of experts for a special session on generative AI...
AMLD 2023 AI workshop, "Set the course: Crafting a legal and regulatory landscape for AI", R. Dubach
Просмотров 1910 месяцев назад
"Set the course: Crafting a legal and regulatory landscape for AI", by Roger Dubach, Ambassador, Deputy Director and Directorate of International Law at EDA - DFAE - FDFA. On Tuesday, August 29th, 2023, to further the discussions from the main AMLD GenAI Conference 2023, the EPFL has brought together another group of experts for a special session on generative AI and AI policy. The realm of AI ...
AMLD 2023 AI workshop, "AI in the machine tool industry: opportunities & challenges", Adina Dorthe
Просмотров 5610 месяцев назад
« AI in the machine tool industry: opportunities & challenges » by Adina Dorthe, Head of Innovation and R&D, Digital Transformation Unit at Georg Fischer. On Tuesday, August 29th, 2023, to further the discussions from the main AMLD GenAI Conference 2023, the EPFL has brought together another group of experts for a special session on generative AI and AI policy. The realm of AI has emerged as a ...
AMLD 2023 AI workshop, « AI Policy - a strategic opportunity », Ayisha Piotti
Просмотров 5310 месяцев назад
« AI Policy - a strategic opportunity », by Ayisha Piotti, Director of AI Policy at ETH Zurich Center for Law and Economics, and Managing Partner at RegHorizon. On Tuesday, August 29th, 2023, to further the discussions from the main AMLD GenAI Conference 2023, the EPFL has brought together another group of experts for a special session on generative AI and AI policy. The realm of AI has emerged...
AMLD 2023 AI workshop round table, "Perspectives from industry, society and research", Aug 29, 2023
Просмотров 7310 месяцев назад
AMLD 2023 AI workshop round table, "Perspectives from industry, society and research", Aug 29, 2023
AMLD 2023 AI workshop, "Welcome introduction", Dr. Jan Kerschgens (EPFL CIS)
Просмотров 5310 месяцев назад
AMLD 2023 AI workshop, "Welcome introduction", Dr. Jan Kerschgens (EPFL CIS)
CIS Colloquium: Prof. Meisam Razaviyayn, University of Southern California
Просмотров 19810 месяцев назад
CIS Colloquium: Prof. Meisam Razaviyayn, University of Southern California
CIS-IMI-C4DT joint event: "AI-driven synthetic realities and trust"
Просмотров 162Год назад
CIS-IMI-C4DT joint event: "AI-driven synthetic realities and trust"
CIS Colloquium: Dr. Petar Veličković, Google DeepMind and University of Cambridge
Просмотров 684Год назад
CIS Colloquium: Dr. Petar Veličković, Google DeepMind and University of Cambridge
CIS Network Seminar: #IDIAP - Prof Sébastien Marcel
Просмотров 85Год назад
CIS Network Seminar: #IDIAP - Prof Sébastien Marcel
Dr. Luiz Chamon: "Learning under requirements"
Просмотров 136Год назад
Dr. Luiz Chamon: "Learning under requirements"
Prof. Michael Herbst: “An interdisciplinary perspective on robust materials simulations”
Просмотров 146Год назад
Prof. Michael Herbst: “An interdisciplinary perspective on robust materials simulations”
Mats Stensrud: “On optimal treatment regimes assisted by algorithms“
Просмотров 131Год назад
Mats Stensrud: “On optimal treatment regimes assisted by algorithms“
Deplancke Lab: Dr. Wouter Saelens - “Mountains in our DNA"
Просмотров 49Год назад
Deplancke Lab: Dr. Wouter Saelens - “Mountains in our DNA"
Gianluca Ceruti (EPFL-CIS): “Low-rank dynamical training of feed-forward neural networks”
Просмотров 157Год назад
Gianluca Ceruti (EPFL-CIS): “Low-rank dynamical training of feed-forward neural networks”
Qibin Zhao (RIKEN-AIP): “Efficient machine learning with tensor networks”
Просмотров 166Год назад
Qibin Zhao (RIKEN-AIP): “Efficient machine learning with tensor networks”
Nicolas Boumal (EPFL-CIS): "Non-convex optimization when the solution is not unique..."
Просмотров 288Год назад
Nicolas Boumal (EPFL-CIS): "Non-convex optimization when the solution is not unique..."
Thomas Moellenhoff (RIKEN-AIP): “SAM as an Optimal Relaxation of Bayes”
Просмотров 245Год назад
Thomas Moellenhoff (RIKEN-AIP): “SAM as an Optimal Relaxation of Bayes”
very good talk, Great
How can costs of digitalbtwins be sustainable
Thank you for great work. Which free parameter tuning method(e.g. Reverse Validation based from DANN paper) do you use?
amazing presentation!
There was no machine learning in this talk.
😩 "promo sm"
Thanks.
Tour de force - great!
If only you people knew what these systems are being used for…. You think it’s to monitor. It isn’t.
So the bottm line is that simple black box method of capturing system dynamics is not the solution. For a digital Twin, you need to apply first principle to model the system and THEN reinforce it with dynamic field data.
Amazing talk
Can someone explain, for prediction tasks, why should we do all of this modeling, instead of building a Deep learning model on the historic data of the physical asset? and retrain it every x amount of time to be up to date. I can understand the interpretability advantage of physics-driven, but is there any other advantage?
yes, spatial interpratation
Because prediction will need training the models? it is a dumb and brute force way to do things. Things would quickly go out of hand. You need lot of computing power.
Promo sm 😻
is it possible to have access to the slides?
Is this related to Canons new SPAD sensor ?
I had that as a first thought in my mind lol. They used this video in that article I saw about the new canon.
Speaker mentioned researcher which went to Canon. He almost on all slides on start of video
One of the excellent and detailed session. I learned many new things. Thanks Respected Professor.
14:55 That made me chuckle a little
This is just an observer?
great work
I am pleased to share with CIS EPFL, with Professionals and with Companies the most advanced technological tool in existence: the Digital Twin that I invented in 1981 under the name of Virtual Instruments. ruclips.net/video/fXlp4QLdWQs/видео.html Export the Revit, Plant 3D, Civil 3D, Inventor file of the Building or Refinery to Excel, prepare Budget 1 and export it to COBRA. Prepare Budget 2 and export it to Microsoft Project. Solve the problems of Overallocated Resources, Planning Problems, prepare the Budget 3 with which the construction of the Building or the Refinery is going to be quoted.
I am pleased to share with CIS EPFL, with Professionals and with Companies the most advanced technological tool in existence: the Digital Twin that I invented in 1981 under the name of Virtual Instruments. ruclips.net/video/fXlp4QLdWQs/видео.html Export the Revit, Plant 3D, Civil 3D, Inventor file of the Building or Refinery to Excel, prepare Budget 1 and export it to COBRA. Prepare Budget 2 and export it to Microsoft Project. Solve the problems of Overallocated Resources, Planning Problems, prepare the Budget 3 with which the construction of the Building or the Refinery is going to be quoted.
I have some questions on this talk I tried to email Prof. Antoine Bosselut but there is no answer!!!
Incredible. Thank you.
Thank you for sharing.
😥 𝐩𝓻Ỗ𝓂Ø𝓈M
thank you so much for this wonderful presentation
The term DT is gaining traction these days in research but I find it repetition of idea used in product/building life cycle. Similarly Reduced Order Model has been there in prototype-model and in phenomenological elements in finite element analysis. I have seen that sometimes, researchers coin new terminology or buzz word that help them to publish their work as new tech.
Vv
Thank you for sharing such a valuable resource
Thank you for your presentation!!!
Awesome video - thanks for sharing 👋
Very inspiring!
My big man congrats 👏 My hero when it came s to computers 🖥 geometric 🙏🏿God bless you sir.
Awesome talk
Man, you cannot come up with an idea in todays world, there will always be someone else. Was looking into something similar and wanted to see whether someone has already done it. Congrats and best wishes @Laurent Daudet
SUCCESS
looking for working under this professor
On hearing the name Prof. Jaggi, one thing quickly pops up in my mind - "Revisiting FW". This work gave a boom in research related to projection-free algorithms..
You are very right man, this man is great man indeed 🖥
"Knowledge of principles is more important than knowledge of certain facts, that's probably a little bit approximate translation." I see what he did there.
Beautiful talk!!! But Mills did not receive Nobel prize. Yang indeed get it, but not for Yang-Mills theory.
"knowledge of principles is more imoprtant than knowledge of certain facts" ... I couldnt quite catch who said that, Elvezios? Does anybody know his name?
2:22