In my talk I will present my work regarding 3D mapping using lidar scanners. I will give an overview of the SLAM problem and its main challenges: robustness, accuracy and processing speed. Regarding robustness and accuracy, we investigate a better point cloud representation based on resampling and surface reconstruction. Moreover, we demonstrate how it can be incorporated in an ICP-based scan matching technique. Finally, we elaborate on globally consistent mapping using loop closures. Regarding processing speed, we propose the integration of our scan matching in a multi-resolution scheme and a GPU-accelerated implementation using our programming language Quasar.
Organizers: Simon Donne
The ability to predict how an environment changes based on forces applied to it is fundamental for a robot to achieve specific goals. Traditionally in robotics, this problem is addressed through the use of pre-specified models or physics simulators, taking advantage of prior knowledge of the problem structure. While these models are general and have broad applicability, they depend on accurate estimation of model parameters such as object shape, mass, friction etc. On the other hand, learning based methods such as Predictive State Representations or more recent deep learning approaches have looked at learning these models directly from raw perceptual information in a model-free manner. These methods operate on raw data without any intermediate parameter estimation, but lack the structure and generality of model-based techniques. In this talk, I will present some work that tries to bridge the gap between these two paradigms by proposing a specific class of deep visual dynamics models (SE3-Nets) that explicitly encode strong physical and 3D geometric priors (specifically, rigid body dynamics) in their structure. As opposed to traditional deep models that reason about dynamics/motion a pixel level, we show that the physical priors implicit in our network architectures enable them to reason about dynamics at the object level - our network learns to identify objects in the scene and to predict rigid body rotation and translation per object. I will present results on applying our deep architectures to two specific problems: 1) Modeling scene dynamics where the task is to predict future depth observations given the current observation and an applied action and 2) Real-time visuomotor control of a Baxter manipulator based only on raw depth data. We show that: 1) Our proposed architectures significantly outperform baseline deep models on dynamics modelling and 2) Our architectures perform comparably or better than baseline models for visuomotor control while operating at camera rates (30Hz) and relying on far less information.
Organizers: Franzi Meier
Machine learning has become a popular application domain for modern optimization techniques, pushing its algorithmic frontier. The need for large scale optimization algorithms which can handle millions of dimensions or data points, typical for the big data era, have brought a resurgence of interest for first order algorithms, making us revisit the venerable stochastic gradient method [Robbins-Monro 1951] as well as the Frank-Wolfe algorithm [Frank-Wolfe 1956]. In this talk, I will review recent improvements on these algorithms which can exploit the structure of modern machine learning approaches. I will explain why the Frank-Wolfe algorithm has become so popular lately; and present a surprising tweak on the stochastic gradient method which yields a fast linear convergence rate. Motivating applications will include weakly supervised video analysis and structured prediction problems.
Organizers: Philipp Hennig
There has been significant prior work on learning realistic, articulated, 3D statistical shape models of the human body. In contrast, there are few such models for animals, despite their many applications in biology, neuroscience, agriculture, and entertainment. The main challenge is that animals are much less cooperative subjects than humans: the best human body models are learned from thousands of 3D scans of people in specific poses, which is infeasible with live animals. In the talk I will illustrate how we extend a state-of-the-art articulated 3D human body model (SMPL) to animals learning from toys a multi-family shape space that can represent lions, cats, dogs, horses, cows and hippos. The generalization of the model is illustrated by fitting it to images of real animals, where it captures realistic animal shapes, even for new species not seen in training.
Moritz Hardt will review some progress and challenges towards preventing discrimination based on sensitive attributes in supervised learning.
In this talk we will give an overview of research efforts within autonomous manipulation at the AASS Research Center, Örebro University, Sweden. We intend to give a holistic view on the historically separated subjects of robot motion planning and control. In particular, viewing motion behavior generation as an optimal control problem allows for a unified formulation that is uncluttered by a-priori domain assumptions and simplified solution strategies. Furthermore, We will also discuss the problems of workspace modeling and perception and how to integrate them in the overarching problem of autonomous manipulation.
Organizers: Ludovic Righetti
As large tensor-variate data increasingly become the norm in applied machine learning and statistics, complex analysis methods similarly increase in prevalence. Such a trend offers the opportunity to understand more intricate features of the data that, ostensibly, could not be studied with simpler datasets or simpler methodologies. While promising, these advances are also perilous: these novel analysis techniques do not always consider the possibility that their results are in fact an expected consequence of some simpler, already-known feature of simpler data (for example, treating the tensor like a matrix or a univariate quantity) or simpler statistic (for example, the mean and covariance of one of the tensor modes). I will present two works that address this growing problem, the first of which uses Kronecker algebra to derive a tensor-variate maximum entropy distribution that shares modal moments with the real data. This distribution of surrogate data forms the basis of a statistical hypothesis test, and I use this method to answer a question of epiphenomenal tensor structure in populations of neural recordings in the motor and prefrontal cortex. In the second part, I will discuss how to extend this maximum entropy formulation to arbitrary constraints using deep neural network architectures in the flavor of implicit generative modeling, and I will use this method in a texture synthesis application.
Organizers: Philipp Hennig
In classical reinforcement learning agents accept arbitrary short term loss for long term gain when exploring their environment. This is infeasible for safety critical applications such as robotics, where even a single unsafe action may cause system failure or harm the environment. In this work, we address the problem of safely exploring finite Markov decision processes (MDP). We define safety in terms of an a priori unknown safety constraint that depends on states and actions and satisfies certain regularity conditions expressed via a Gaussian process prior. We develop a novel algorithm, SAFEMDP, for this task and prove that it completely explores the safely reachable part of the MDP without violating the safety constraint. Moreover, the algorithm explicitly considers reachability when exploring the MDP, ensuring that it does not get stuck in any state with no safe way out. We demonstrate our method on digital terrain models for the task of exploring an unknown map with a rover.
Organizers: Sebastian Trimpe
Organizers: Moritz Grosse-Wentrup
This is the story of the novel model predictive control (MPC) solution for ABB’s largest drive, the Megadrive LCI. LCI stands for load commutated inverter, a type of current source converter which powers large machineries in many industries such as marine, mining or oil & gas. Starting from a small software project at ABB Corporate Research, this novel control solution turned out to become the first time ever MPC was employed in a 48 MW commercial drive. Subsequently it was commissioned at Kollsnes, a key facility of the natural gas delivery chain, in order to increase the plant’s availability. In this presentation I will talk about the magic behind this success story, the so-called Embedded MPC algorithms, and my objective will be to demonstrate the possibilities when power meets computation.
Organizers: Sebastian Trimpe
Brain-Computer Interfaces (BCIs) are systems that can translate brain activity patterns of a user into messages or commands for an interactive application. Such brain activity is typically measured using Electroencephalography (EEG), before being processed and classified by the system. EEG-based BCIs have proven promising for a wide range of applications ranging from communication and control for motor impaired users, to gaming targeted at the general public, real-time mental state monitoring and stroke rehabilitation, to name a few. Despite this promising potential, BCIs are still scarcely used outside laboratories for practical applications. The main reason preventing EEG-based BCIs from being widely used is arguably their poor usability, which is notably due to their low robustness and reliability, as well as their long training times. In this talk I present some of our research aimed at addressing these points in order to make EEG-based BCIs usable, i.e., to increase their efficacy and efficiency. In particular, I will present a set of contributions towards this goal 1) at the user training level, to ensure that users can learn to control a BCI efficiently and effectively, and 2) at the usage level, to explore novel applications of BCIs for which the current reliability can already be useful, e.g., for neuroergonomics or real-time brain activity and mental state visualization.
The predictive simulation of engineering systems increasingly rests on the synthesis of physical models and experimental data. In this context, Bayesian inference establishes a framework for quantifying the encountered uncertainties and fusing the available information. A summary and discussion of some recently emerged methods for uncertainty propagation (polynomial chaos expansions) and related MCMC-free techniques for posterior computation (spectral likelihood expansions, optimal transportation theory) is presented.
Organizers: Philipp Hennig
In this talk I am going to present the work we have been doing at the Computer Vision Lab of the Technical University of Munich which started as an attempt to better deal with videos (and therefore the time domain) within neural network architectures.
Organizers: Joel Janai