Project acronym 3D Reloaded
Project 3D Reloaded: Novel Algorithms for 3D Shape Inference and Analysis
Researcher (PI) Daniel Cremers
Host Institution (HI) TECHNISCHE UNIVERSITAET MUENCHEN
Call Details Consolidator Grant (CoG), PE6, ERC-2014-CoG
Summary Despite their amazing success, we believe that computer vision algorithms have only scratched the surface of what can be done in terms of modeling and understanding our world from images. We believe that novel image analysis techniques will be a major enabler and driving force behind next-generation technologies, enhancing everyday life and opening up radically new possibilities. And we believe that the key to achieving this is to develop algorithms for reconstructing and analyzing the 3D structure of our world.
In this project, we will focus on three lines of research:
A) We will develop algorithms for 3D reconstruction from standard color cameras and from RGB-D cameras. In particular, we will promote real-time-capable direct and dense methods. In contrast to the classical two-stage approach of sparse feature-point based motion estimation and subsequent dense reconstruction, these methods optimally exploit all color information to jointly estimate dense geometry and camera motion.
B) We will develop algorithms for 3D shape analysis, including rigid and non-rigid matching, decomposition and interpretation of 3D shapes. We will focus on algorithms which are optimal or near-optimal. One of the major computational challenges lies in generalizing existing 2D shape analysis techniques to shapes in 3D and 4D (temporal evolutions of 3D shape).
C) We will develop shape priors for 3D reconstruction. These can be learned from sample shapes or acquired during the reconstruction process. For example, when reconstructing a larger office algorithms may exploit the geometric self-similarity of the scene, storing a model of a chair and its multiple instances only once rather than multiple times.
Advancing the state of the art in geometric reconstruction and geometric analysis will have a profound impact well beyond computer vision. We strongly believe that we have the necessary competence to pursue this project. Preliminary results have been well received by the community.
Summary
Despite their amazing success, we believe that computer vision algorithms have only scratched the surface of what can be done in terms of modeling and understanding our world from images. We believe that novel image analysis techniques will be a major enabler and driving force behind next-generation technologies, enhancing everyday life and opening up radically new possibilities. And we believe that the key to achieving this is to develop algorithms for reconstructing and analyzing the 3D structure of our world.
In this project, we will focus on three lines of research:
A) We will develop algorithms for 3D reconstruction from standard color cameras and from RGB-D cameras. In particular, we will promote real-time-capable direct and dense methods. In contrast to the classical two-stage approach of sparse feature-point based motion estimation and subsequent dense reconstruction, these methods optimally exploit all color information to jointly estimate dense geometry and camera motion.
B) We will develop algorithms for 3D shape analysis, including rigid and non-rigid matching, decomposition and interpretation of 3D shapes. We will focus on algorithms which are optimal or near-optimal. One of the major computational challenges lies in generalizing existing 2D shape analysis techniques to shapes in 3D and 4D (temporal evolutions of 3D shape).
C) We will develop shape priors for 3D reconstruction. These can be learned from sample shapes or acquired during the reconstruction process. For example, when reconstructing a larger office algorithms may exploit the geometric self-similarity of the scene, storing a model of a chair and its multiple instances only once rather than multiple times.
Advancing the state of the art in geometric reconstruction and geometric analysis will have a profound impact well beyond computer vision. We strongly believe that we have the necessary competence to pursue this project. Preliminary results have been well received by the community.
Max ERC Funding
2 000 000 €
Duration
Start date: 2015-09-01, End date: 2020-08-31
Project acronym 4DRepLy
Project Closing the 4D Real World Reconstruction Loop
Researcher (PI) Christian THEOBALT
Host Institution (HI) MAX-PLANCK-GESELLSCHAFT ZUR FORDERUNG DER WISSENSCHAFTEN EV
Call Details Consolidator Grant (CoG), PE6, ERC-2017-COG
Summary 4D reconstruction, the camera-based dense dynamic scene reconstruction, is a grand challenge in computer graphics and computer vision. Despite great progress, 4D capturing the complex, diverse real world outside a studio is still far from feasible. 4DRepLy builds a new generation of high-fidelity 4D reconstruction (4DRecon) methods. They will be the first to efficiently capture all types of deformable objects (humans and other types) in crowded real world scenes with a single color or depth camera. They capture space-time coherent deforming geometry, motion, high-frequency reflectance and illumination at unprecedented detail, and will be the first to handle difficult occlusions, topology changes and large groups of interacting objects. They automatically adapt to new scene types, yet deliver models with meaningful, interpretable parameters. This requires far reaching contributions: First, we develop groundbreaking new plasticity-enhanced model-based 4D reconstruction methods that automatically adapt to new scenes. Second, we develop radically new machine learning-based dense 4D reconstruction methods. Third, these model- and learning-based methods are combined in two revolutionary new classes of 4DRecon methods: 1) advanced fusion-based methods and 2) methods with deep architectural integration. Both, 1) and 2), are automatically designed in the 4D Real World Reconstruction Loop, a revolutionary new design paradigm in which 4DRecon methods refine and adapt themselves while continuously processing unlabeled real world input. This overcomes the previously unbreakable scalability barrier to real world scene diversity, complexity and generality. This paradigm shift opens up a new research direction in graphics and vision and has far reaching relevance across many scientific fields. It enables new applications of profound social pervasion and significant economic impact, e.g., for visual media and virtual/augmented reality, and for future autonomous and robotic systems.
Summary
4D reconstruction, the camera-based dense dynamic scene reconstruction, is a grand challenge in computer graphics and computer vision. Despite great progress, 4D capturing the complex, diverse real world outside a studio is still far from feasible. 4DRepLy builds a new generation of high-fidelity 4D reconstruction (4DRecon) methods. They will be the first to efficiently capture all types of deformable objects (humans and other types) in crowded real world scenes with a single color or depth camera. They capture space-time coherent deforming geometry, motion, high-frequency reflectance and illumination at unprecedented detail, and will be the first to handle difficult occlusions, topology changes and large groups of interacting objects. They automatically adapt to new scene types, yet deliver models with meaningful, interpretable parameters. This requires far reaching contributions: First, we develop groundbreaking new plasticity-enhanced model-based 4D reconstruction methods that automatically adapt to new scenes. Second, we develop radically new machine learning-based dense 4D reconstruction methods. Third, these model- and learning-based methods are combined in two revolutionary new classes of 4DRecon methods: 1) advanced fusion-based methods and 2) methods with deep architectural integration. Both, 1) and 2), are automatically designed in the 4D Real World Reconstruction Loop, a revolutionary new design paradigm in which 4DRecon methods refine and adapt themselves while continuously processing unlabeled real world input. This overcomes the previously unbreakable scalability barrier to real world scene diversity, complexity and generality. This paradigm shift opens up a new research direction in graphics and vision and has far reaching relevance across many scientific fields. It enables new applications of profound social pervasion and significant economic impact, e.g., for visual media and virtual/augmented reality, and for future autonomous and robotic systems.
Max ERC Funding
1 977 000 €
Duration
Start date: 2018-09-01, End date: 2023-08-31
Project acronym ACCORD
Project Algorithms for Complex Collective Decisions on Structured Domains
Researcher (PI) Edith Elkind
Host Institution (HI) THE CHANCELLOR, MASTERS AND SCHOLARS OF THE UNIVERSITY OF OXFORD
Call Details Starting Grant (StG), PE6, ERC-2014-STG
Summary Algorithms for Complex Collective Decisions on Structured Domains.
The aim of this proposal is to substantially advance the field of Computational Social Choice, by developing new tools and methodologies that can be used for making complex group decisions in rich and structured environments. We consider settings where each member of a decision-making body has preferences over a finite set of alternatives, and the goal is to synthesise a collective preference over these alternatives, which may take the form of a partial order over the set of alternatives with a predefined structure: examples include selecting a fixed-size set of alternatives, a ranking of the alternatives, a winner and up to two runner-ups, etc. We will formulate desiderata that apply to such preference aggregation procedures, design specific procedures that satisfy as many of these desiderata as possible, and develop efficient algorithms for computing them. As the latter step may be infeasible on general preference domains, we will focus on identifying the least restrictive domains that enable efficient computation, and use real-life preference data to verify whether the associated restrictions are likely to be satisfied in realistic preference aggregation scenarios. Also, we will determine whether our preference aggregation procedures are computationally resistant to malicious behavior. To lower the cognitive burden on the decision-makers, we will extend our procedures to accept partial rankings as inputs. Finally, to further contribute towards bridging the gap between theory and practice of collective decision making, we will provide open-source software implementations of our procedures, and reach out to the potential users to obtain feedback on their practical applicability.
Summary
Algorithms for Complex Collective Decisions on Structured Domains.
The aim of this proposal is to substantially advance the field of Computational Social Choice, by developing new tools and methodologies that can be used for making complex group decisions in rich and structured environments. We consider settings where each member of a decision-making body has preferences over a finite set of alternatives, and the goal is to synthesise a collective preference over these alternatives, which may take the form of a partial order over the set of alternatives with a predefined structure: examples include selecting a fixed-size set of alternatives, a ranking of the alternatives, a winner and up to two runner-ups, etc. We will formulate desiderata that apply to such preference aggregation procedures, design specific procedures that satisfy as many of these desiderata as possible, and develop efficient algorithms for computing them. As the latter step may be infeasible on general preference domains, we will focus on identifying the least restrictive domains that enable efficient computation, and use real-life preference data to verify whether the associated restrictions are likely to be satisfied in realistic preference aggregation scenarios. Also, we will determine whether our preference aggregation procedures are computationally resistant to malicious behavior. To lower the cognitive burden on the decision-makers, we will extend our procedures to accept partial rankings as inputs. Finally, to further contribute towards bridging the gap between theory and practice of collective decision making, we will provide open-source software implementations of our procedures, and reach out to the potential users to obtain feedback on their practical applicability.
Max ERC Funding
1 395 933 €
Duration
Start date: 2015-07-01, End date: 2020-06-30
Project acronym ACDC
Project Algorithms and Complexity of Highly Decentralized Computations
Researcher (PI) Fabian Daniel Kuhn
Host Institution (HI) ALBERT-LUDWIGS-UNIVERSITAET FREIBURG
Call Details Starting Grant (StG), PE6, ERC-2013-StG
Summary "Many of today's and tomorrow's computer systems are built on top of large-scale networks such as, e.g., the Internet, the world wide web, wireless ad hoc and sensor networks, or peer-to-peer networks. Driven by technological advances, new kinds of networks and applications have become possible and we can safely assume that this trend is going to continue. Often modern systems are envisioned to consist of a potentially large number of individual components that are organized in a completely decentralized way. There is no central authority that controls the topology of the network, how nodes join or leave the system, or in which way nodes communicate with each other. Also, many future distributed applications will be built using wireless devices that communicate via radio.
The general objective of the proposed project is to improve our understanding of the algorithmic and theoretical foundations of decentralized distributed systems. From an algorithmic point of view, decentralized networks and computations pose a number of fascinating and unique challenges that are not present in sequential or more standard distributed systems. As communication is limited and mostly between nearby nodes, each node of a large network can only maintain a very restricted view of the global state of the system. This is particularly true if the network can change dynamically, either by nodes joining or leaving the system or if the topology changes over time, e.g., because of the mobility of the devices in case of a wireless network. Nevertheless, the nodes of a network need to coordinate in order to achieve some global goal.
In particular, we plan to study algorithms and lower bounds for basic computation and information dissemination tasks in such systems. In addition, we are particularly interested in the complexity of distributed computations in dynamic and wireless networks."
Summary
"Many of today's and tomorrow's computer systems are built on top of large-scale networks such as, e.g., the Internet, the world wide web, wireless ad hoc and sensor networks, or peer-to-peer networks. Driven by technological advances, new kinds of networks and applications have become possible and we can safely assume that this trend is going to continue. Often modern systems are envisioned to consist of a potentially large number of individual components that are organized in a completely decentralized way. There is no central authority that controls the topology of the network, how nodes join or leave the system, or in which way nodes communicate with each other. Also, many future distributed applications will be built using wireless devices that communicate via radio.
The general objective of the proposed project is to improve our understanding of the algorithmic and theoretical foundations of decentralized distributed systems. From an algorithmic point of view, decentralized networks and computations pose a number of fascinating and unique challenges that are not present in sequential or more standard distributed systems. As communication is limited and mostly between nearby nodes, each node of a large network can only maintain a very restricted view of the global state of the system. This is particularly true if the network can change dynamically, either by nodes joining or leaving the system or if the topology changes over time, e.g., because of the mobility of the devices in case of a wireless network. Nevertheless, the nodes of a network need to coordinate in order to achieve some global goal.
In particular, we plan to study algorithms and lower bounds for basic computation and information dissemination tasks in such systems. In addition, we are particularly interested in the complexity of distributed computations in dynamic and wireless networks."
Max ERC Funding
1 148 000 €
Duration
Start date: 2013-11-01, End date: 2018-10-31
Project acronym ACROSS
Project 3D Reconstruction and Modeling across Different Levels of Abstraction
Researcher (PI) Leif Kobbelt
Host Institution (HI) RHEINISCH-WESTFAELISCHE TECHNISCHE HOCHSCHULE AACHEN
Call Details Advanced Grant (AdG), PE6, ERC-2013-ADG
Summary "Digital 3D models are gaining more and more importance in diverse application fields ranging from computer graphics, multimedia and simulation sciences to engineering, architecture, and medicine. Powerful technologies to digitize the 3D shape of real objects and scenes are becoming available even to consumers. However, the raw geometric data emerging from, e.g., 3D scanning or multi-view stereo often lacks a consistent structure and meta-information which are necessary for the effective deployment of such models in sophisticated down-stream applications like animation, simulation, or CAD/CAM that go beyond mere visualization. Our goal is to develop new fundamental algorithms which transform raw geometric input data into augmented 3D models that are equipped with structural meta information such as feature aligned meshes, patch segmentations, local and global geometric constraints, statistical shape variation data, or even procedural descriptions. Our methodological approach is inspired by the human perceptual system that integrates bottom-up (data-driven) and top-down (model-driven) mechanisms in its hierarchical processing. Similarly we combine algorithms operating on different levels of abstraction into reconstruction and modeling networks. Instead of developing an individual solution for each specific application scenario, we create an eco-system of algorithms for automatic processing and interactive design of highly complex 3D models. A key concept is the information flow across all levels of abstraction in a bottom-up as well as top-down fashion. We not only aim at optimizing geometric representations but in fact at bridging the gap between reconstruction and recognition of geometric objects. The results from this project will make it possible to bring 3D models of real world objects into many highly relevant applications in science, industry, and entertainment, greatly reducing the excessive manual effort that is still necessary today."
Summary
"Digital 3D models are gaining more and more importance in diverse application fields ranging from computer graphics, multimedia and simulation sciences to engineering, architecture, and medicine. Powerful technologies to digitize the 3D shape of real objects and scenes are becoming available even to consumers. However, the raw geometric data emerging from, e.g., 3D scanning or multi-view stereo often lacks a consistent structure and meta-information which are necessary for the effective deployment of such models in sophisticated down-stream applications like animation, simulation, or CAD/CAM that go beyond mere visualization. Our goal is to develop new fundamental algorithms which transform raw geometric input data into augmented 3D models that are equipped with structural meta information such as feature aligned meshes, patch segmentations, local and global geometric constraints, statistical shape variation data, or even procedural descriptions. Our methodological approach is inspired by the human perceptual system that integrates bottom-up (data-driven) and top-down (model-driven) mechanisms in its hierarchical processing. Similarly we combine algorithms operating on different levels of abstraction into reconstruction and modeling networks. Instead of developing an individual solution for each specific application scenario, we create an eco-system of algorithms for automatic processing and interactive design of highly complex 3D models. A key concept is the information flow across all levels of abstraction in a bottom-up as well as top-down fashion. We not only aim at optimizing geometric representations but in fact at bridging the gap between reconstruction and recognition of geometric objects. The results from this project will make it possible to bring 3D models of real world objects into many highly relevant applications in science, industry, and entertainment, greatly reducing the excessive manual effort that is still necessary today."
Max ERC Funding
2 482 000 €
Duration
Start date: 2014-03-01, End date: 2019-02-28
Project acronym Active-DNA
Project Computationally Active DNA Nanostructures
Researcher (PI) Damien WOODS
Host Institution (HI) NATIONAL UNIVERSITY OF IRELAND MAYNOOTH
Call Details Consolidator Grant (CoG), PE6, ERC-2017-COG
Summary During the 20th century computer technology evolved from bulky, slow, special purpose mechanical engines to the now ubiquitous silicon chips and software that are one of the pinnacles of human ingenuity. The goal of the field of molecular programming is to take the next leap and build a new generation of matter-based computers using DNA, RNA and proteins. This will be accomplished by computer scientists, physicists and chemists designing molecules to execute ``wet'' nanoscale programs in test tubes. The workflow includes proposing theoretical models, mathematically proving their computational properties, physical modelling and implementation in the wet-lab.
The past decade has seen remarkable progress at building static 2D and 3D DNA nanostructures. However, unlike biological macromolecules and complexes that are built via specified self-assembly pathways, that execute robotic-like movements, and that undergo evolution, the activity of human-engineered nanostructures is severely limited. We will need sophisticated algorithmic ideas to build structures that rival active living systems. Active-DNA, aims to address this challenge by achieving a number of objectives on computation, DNA-based self-assembly and molecular robotics. Active-DNA research work will range from defining models and proving theorems that characterise the computational and expressive capabilities of such active programmable materials to experimental work implementing active DNA nanostructures in the wet-lab.
Summary
During the 20th century computer technology evolved from bulky, slow, special purpose mechanical engines to the now ubiquitous silicon chips and software that are one of the pinnacles of human ingenuity. The goal of the field of molecular programming is to take the next leap and build a new generation of matter-based computers using DNA, RNA and proteins. This will be accomplished by computer scientists, physicists and chemists designing molecules to execute ``wet'' nanoscale programs in test tubes. The workflow includes proposing theoretical models, mathematically proving their computational properties, physical modelling and implementation in the wet-lab.
The past decade has seen remarkable progress at building static 2D and 3D DNA nanostructures. However, unlike biological macromolecules and complexes that are built via specified self-assembly pathways, that execute robotic-like movements, and that undergo evolution, the activity of human-engineered nanostructures is severely limited. We will need sophisticated algorithmic ideas to build structures that rival active living systems. Active-DNA, aims to address this challenge by achieving a number of objectives on computation, DNA-based self-assembly and molecular robotics. Active-DNA research work will range from defining models and proving theorems that characterise the computational and expressive capabilities of such active programmable materials to experimental work implementing active DNA nanostructures in the wet-lab.
Max ERC Funding
2 349 603 €
Duration
Start date: 2018-11-01, End date: 2023-10-31
Project acronym ACTIVIA
Project Visual Recognition of Function and Intention
Researcher (PI) Ivan Laptev
Host Institution (HI) INSTITUT NATIONAL DE RECHERCHE ENINFORMATIQUE ET AUTOMATIQUE
Call Details Starting Grant (StG), PE6, ERC-2012-StG_20111012
Summary "Computer vision is concerned with the automated interpretation of images and video streams. Today's research is (mostly) aimed at answering queries such as ""Is this a picture of a dog?"", (classification) or sometimes ""Find the dog in this photo"" (detection). While categorisation and detection are useful for many tasks, inferring correct class labels is not the final answer to visual recognition. The categories and locations of objects do not provide direct understanding of their function i.e., how things work, what they can be used for, or how they can act and react. Such an understanding, however, would be highly desirable to answer currently unsolvable queries such as ""Am I in danger?"" or ""What can happen in this scene?"". Solving such queries is the aim of this proposal.
My goal is to uncover the functional properties of objects and the purpose of actions by addressing visual recognition from a different and yet unexplored perspective. The main novelty of this proposal is to leverage observations of people, i.e., their actions and interactions to automatically learn the use, the purpose and the function of objects and scenes from visual data. The project is timely as it builds upon the two key recent technological advances: (a) the immense progress in visual recognition of objects, scenes and human actions achieved in the last ten years, as well as (b) the emergence of a massive amount of public image and video data now available to train visual models.
ACTIVIA addresses fundamental research issues in automated interpretation of dynamic visual scenes, but its results are expected to serve as a basis for ground-breaking technological advances in practical applications. The recognition of functional properties and intentions as explored in this project will directly support high-impact applications such as detection of abnormal events, which are likely to revolutionise today's approaches to crime protection, hazard prevention, elderly care, and many others."
Summary
"Computer vision is concerned with the automated interpretation of images and video streams. Today's research is (mostly) aimed at answering queries such as ""Is this a picture of a dog?"", (classification) or sometimes ""Find the dog in this photo"" (detection). While categorisation and detection are useful for many tasks, inferring correct class labels is not the final answer to visual recognition. The categories and locations of objects do not provide direct understanding of their function i.e., how things work, what they can be used for, or how they can act and react. Such an understanding, however, would be highly desirable to answer currently unsolvable queries such as ""Am I in danger?"" or ""What can happen in this scene?"". Solving such queries is the aim of this proposal.
My goal is to uncover the functional properties of objects and the purpose of actions by addressing visual recognition from a different and yet unexplored perspective. The main novelty of this proposal is to leverage observations of people, i.e., their actions and interactions to automatically learn the use, the purpose and the function of objects and scenes from visual data. The project is timely as it builds upon the two key recent technological advances: (a) the immense progress in visual recognition of objects, scenes and human actions achieved in the last ten years, as well as (b) the emergence of a massive amount of public image and video data now available to train visual models.
ACTIVIA addresses fundamental research issues in automated interpretation of dynamic visual scenes, but its results are expected to serve as a basis for ground-breaking technological advances in practical applications. The recognition of functional properties and intentions as explored in this project will directly support high-impact applications such as detection of abnormal events, which are likely to revolutionise today's approaches to crime protection, hazard prevention, elderly care, and many others."
Max ERC Funding
1 497 420 €
Duration
Start date: 2013-01-01, End date: 2018-12-31
Project acronym ACUITY
Project Algorithms for coping with uncertainty and intractability
Researcher (PI) Nikhil Bansal
Host Institution (HI) TECHNISCHE UNIVERSITEIT EINDHOVEN
Call Details Consolidator Grant (CoG), PE6, ERC-2013-CoG
Summary The two biggest challenges in solving practical optimization problems are computational intractability, and the presence
of uncertainty: most problems are either NP-hard, or have incomplete input data which
makes an exact computation impossible.
Recently, there has been a huge progress in our understanding of intractability, based on spectacular algorithmic and lower bound techniques. For several problems, especially those with only local constraints, we can design optimum
approximation algorithms that are provably the best possible.
However, typical optimization problems usually involve complex global constraints and are much less understood. The situation is even worse for coping with uncertainty. Most of the algorithms are based on ad-hoc techniques and there is no deeper understanding of what makes various problems easy or hard.
This proposal describes several new directions, together with concrete intermediate goals, that will break important new ground in the theory of approximation and online algorithms. The particular directions we consider are (i) extend the primal dual method to systematically design online algorithms, (ii) build a structural theory of online problems based on work functions, (iii) develop new tools to use the power of strong convex relaxations and (iv) design new algorithmic approaches based on non-constructive proof techniques.
The proposed research is at the
cutting edge of algorithm design, and builds upon the recent success of the PI in resolving several longstanding questions in these areas. Any progress is likely to be a significant contribution to theoretical
computer science and combinatorial optimization.
Summary
The two biggest challenges in solving practical optimization problems are computational intractability, and the presence
of uncertainty: most problems are either NP-hard, or have incomplete input data which
makes an exact computation impossible.
Recently, there has been a huge progress in our understanding of intractability, based on spectacular algorithmic and lower bound techniques. For several problems, especially those with only local constraints, we can design optimum
approximation algorithms that are provably the best possible.
However, typical optimization problems usually involve complex global constraints and are much less understood. The situation is even worse for coping with uncertainty. Most of the algorithms are based on ad-hoc techniques and there is no deeper understanding of what makes various problems easy or hard.
This proposal describes several new directions, together with concrete intermediate goals, that will break important new ground in the theory of approximation and online algorithms. The particular directions we consider are (i) extend the primal dual method to systematically design online algorithms, (ii) build a structural theory of online problems based on work functions, (iii) develop new tools to use the power of strong convex relaxations and (iv) design new algorithmic approaches based on non-constructive proof techniques.
The proposed research is at the
cutting edge of algorithm design, and builds upon the recent success of the PI in resolving several longstanding questions in these areas. Any progress is likely to be a significant contribution to theoretical
computer science and combinatorial optimization.
Max ERC Funding
1 519 285 €
Duration
Start date: 2014-05-01, End date: 2019-04-30
Project acronym ADAPT
Project Theory and Algorithms for Adaptive Particle Simulation
Researcher (PI) Stephane Redon
Host Institution (HI) INSTITUT NATIONAL DE RECHERCHE ENINFORMATIQUE ET AUTOMATIQUE
Call Details Starting Grant (StG), PE6, ERC-2012-StG_20111012
Summary "During the twentieth century, the development of macroscopic engineering has been largely stimulated by progress in digital prototyping: cars, planes, boats, etc. are nowadays designed and tested on computers. Digital prototypes have progressively replaced actual ones, and effective computer-aided engineering tools have helped cut costs and reduce production cycles of these macroscopic systems.
The twenty-first century is most likely to see a similar development at the atomic scale. Indeed, the recent years have seen tremendous progress in nanotechnology - in particular in the ability to control matter at the atomic scale. Similar to what has happened with macroscopic engineering, powerful and generic computational tools will be needed to engineer complex nanosystems, through modeling and simulation. As a result, a major challenge is to develop efficient simulation methods and algorithms.
NANO-D, the INRIA research group I started in January 2008 in Grenoble, France, aims at developing
efficient computational methods for modeling and simulating complex nanosystems, both natural and artificial. In particular, NANO-D develops SAMSON, a software application which gathers all algorithms designed by the group and its collaborators (SAMSON: Software for Adaptive Modeling and Simulation Of Nanosystems).
In this project, I propose to develop a unified theory, and associated algorithms, for adaptive particle simulation. The proposed theory will avoid problems that plague current popular multi-scale or hybrid simulation approaches by simulating a single potential throughout the system, while allowing users to finely trade precision for computational speed.
I believe the full development of the adaptive particle simulation theory will have an important impact on current modeling and simulation practices, and will enable practical design of complex nanosystems on desktop computers, which should significantly boost the emergence of generic nano-engineering."
Summary
"During the twentieth century, the development of macroscopic engineering has been largely stimulated by progress in digital prototyping: cars, planes, boats, etc. are nowadays designed and tested on computers. Digital prototypes have progressively replaced actual ones, and effective computer-aided engineering tools have helped cut costs and reduce production cycles of these macroscopic systems.
The twenty-first century is most likely to see a similar development at the atomic scale. Indeed, the recent years have seen tremendous progress in nanotechnology - in particular in the ability to control matter at the atomic scale. Similar to what has happened with macroscopic engineering, powerful and generic computational tools will be needed to engineer complex nanosystems, through modeling and simulation. As a result, a major challenge is to develop efficient simulation methods and algorithms.
NANO-D, the INRIA research group I started in January 2008 in Grenoble, France, aims at developing
efficient computational methods for modeling and simulating complex nanosystems, both natural and artificial. In particular, NANO-D develops SAMSON, a software application which gathers all algorithms designed by the group and its collaborators (SAMSON: Software for Adaptive Modeling and Simulation Of Nanosystems).
In this project, I propose to develop a unified theory, and associated algorithms, for adaptive particle simulation. The proposed theory will avoid problems that plague current popular multi-scale or hybrid simulation approaches by simulating a single potential throughout the system, while allowing users to finely trade precision for computational speed.
I believe the full development of the adaptive particle simulation theory will have an important impact on current modeling and simulation practices, and will enable practical design of complex nanosystems on desktop computers, which should significantly boost the emergence of generic nano-engineering."
Max ERC Funding
1 476 882 €
Duration
Start date: 2012-09-01, End date: 2017-08-31
Project acronym AI4REASON
Project Artificial Intelligence for Large-Scale Computer-Assisted Reasoning
Researcher (PI) Josef Urban
Host Institution (HI) CESKE VYSOKE UCENI TECHNICKE V PRAZE
Call Details Consolidator Grant (CoG), PE6, ERC-2014-CoG
Summary The goal of the AI4REASON project is a breakthrough in what is considered a very hard problem in AI and automation of reasoning, namely the problem of automatically proving theorems in large and complex theories. Such complex formal theories arise in projects aimed at verification of today's advanced mathematics such as the Formal Proof of the Kepler Conjecture (Flyspeck), verification of software and hardware designs such as the seL4 operating system kernel, and verification of other advanced systems and technologies on which today's information society critically depends.
It seems extremely complex and unlikely to design an explicitly programmed solution to the problem. However, we have recently demonstrated that the performance of existing approaches can be multiplied by data-driven AI methods that learn reasoning guidance from large proof corpora. The breakthrough will be achieved by developing such novel AI methods. First, we will devise suitable Automated Reasoning and Machine Learning methods that learn reasoning knowledge and steer the reasoning processes at various levels of granularity. Second, we will combine them into autonomous self-improving AI systems that interleave deduction and learning in positive feedback loops. Third, we will develop approaches that aggregate reasoning knowledge across many formal, semi-formal and informal corpora and deploy the methods as strong automation services for the formal proof community.
The expected outcome is our ability to prove automatically at least 50% more theorems in high-assurance projects such as Flyspeck and seL4, bringing a major breakthrough in formal reasoning and verification. As an AI effort, the project offers a unique path to large-scale semantic AI. The formal corpora concentrate centuries of deep human thinking in a computer-understandable form on which deductive and inductive AI can be combined and co-evolved, providing new insights into how humans do mathematics and science.
Summary
The goal of the AI4REASON project is a breakthrough in what is considered a very hard problem in AI and automation of reasoning, namely the problem of automatically proving theorems in large and complex theories. Such complex formal theories arise in projects aimed at verification of today's advanced mathematics such as the Formal Proof of the Kepler Conjecture (Flyspeck), verification of software and hardware designs such as the seL4 operating system kernel, and verification of other advanced systems and technologies on which today's information society critically depends.
It seems extremely complex and unlikely to design an explicitly programmed solution to the problem. However, we have recently demonstrated that the performance of existing approaches can be multiplied by data-driven AI methods that learn reasoning guidance from large proof corpora. The breakthrough will be achieved by developing such novel AI methods. First, we will devise suitable Automated Reasoning and Machine Learning methods that learn reasoning knowledge and steer the reasoning processes at various levels of granularity. Second, we will combine them into autonomous self-improving AI systems that interleave deduction and learning in positive feedback loops. Third, we will develop approaches that aggregate reasoning knowledge across many formal, semi-formal and informal corpora and deploy the methods as strong automation services for the formal proof community.
The expected outcome is our ability to prove automatically at least 50% more theorems in high-assurance projects such as Flyspeck and seL4, bringing a major breakthrough in formal reasoning and verification. As an AI effort, the project offers a unique path to large-scale semantic AI. The formal corpora concentrate centuries of deep human thinking in a computer-understandable form on which deductive and inductive AI can be combined and co-evolved, providing new insights into how humans do mathematics and science.
Max ERC Funding
1 499 500 €
Duration
Start date: 2015-09-01, End date: 2020-08-31
