arXiv subjects

Matthias Müller

Publications and source records attributed to Matthias Müller.

46 records · Page 3Linked to original sources

arXiv2019

Resonant X-ray Emission and Valence-band Lifetime Broadening in LiNO$_3$

X-ray absorption and resonant inelastic x-ray scattering measurements are carried out on lithium nitrate LiNO$_3$. The $σ$ orbitals around the nitrogen atoms exhibit a large lifetime effect. Experimentally, this is manifest as an apparent weakening of the x-ray emission signal from these states, but a closer examination shows that instead it is due to extreme broadening. This echos previous studies on ammonium nitrate, which, despite large differences in the cation and space group, showed a similar effect associated with the nitrate. Using first-principles $GW$ self-energy and Bethe-Salpeter equation calculations we show that this effect is due in part to short quasi-hole lifetimes for the orbitals constituting the NO $σ$ bonds.

cond-mat.mtrl-sci↗

arXiv2019

Learning a Controller Fusion Network by Online Trajectory Filtering for Vision-based UAV Racing

Autonomous UAV racing has recently emerged as an interesting research problem. The dream is to beat humans in this new fast-paced sport. A common approach is to learn an end-to-end policy that directly predicts controls from raw images by imitating an expert. However, such a policy is limited by the expert it imitates and scaling to other environments and vehicle dynamics is difficult. One approach to overcome the drawbacks of an end-to-end policy is to train a network only on the perception task and handle control with a PID or MPC controller. However, a single controller must be extensively tuned and cannot usually cover the whole state space. In this paper, we propose learning an optimized controller using a DNN that fuses multiple controllers. The network learns a robust controller with online trajectory filtering, which suppresses noisy trajectories and imperfections of individual controllers. The result is a network that is able to learn a good fusion of filtered trajectories from different controllers leading to significant improvements in overall performance. We compare our trained network to controllers it has learned from, end-to-end baselines and human pilots in a realistic simulation; our network beats all baselines in extensive experiments and approaches the performance of a professional human pilot. A video summarizing this work is available at https://youtu.be/hGKlE5X9Z5U

cs.RO↗

arXiv2019

Operation of a Microfabricated Planar Ion-Trap for Studies of a Yb$^+$-Rb Hybrid Quantum System

In order to study interactions of atomic ions with ultracold neutral atoms, it is important to have sub-$μ$m control over positioning ion crystals. Serving for this purpose, we introduce a microfabricated planar ion trap featuring 21 DC electrodes. The ion trap is controlled by a home-made FPGA voltage source providing independently variable voltages to each of the DC electrodes. To assure stable positioning of ion crystals with respect to trapped neutral atoms, we integrate into the overall design a compact mirror magneto optical chip trap (mMOT) for cooling and confining neutral $^{87}$Rb atoms. The trapped atoms will be transferred into an also integrated chipbased Ioffe-Pritchard trap potential formed by a Z-shaped wire and an external bias magnetic field.We introduce the hybrid atom-ion chip, the microfabricated planar ion trap and use trapped ion crystals to determine ion lifetimes, trap frequencies, positioning ions and the accuracy of the compensation of micromotion.

physics.atom-ph↗

arXiv2018

Driving Policy Transfer via Modularity and Abstraction

End-to-end approaches to autonomous driving have high sample complexity and are difficult to scale to realistic urban driving. Simulation can help end-to-end driving systems by providing a cheap, safe, and diverse training environment. Yet training driving policies in simulation brings up the problem of transferring such policies to the real world. We present an approach to transferring driving policies from simulation to reality via modularity and abstraction. Our approach is inspired by classic driving systems and aims to combine the benefits of modular architectures and end-to-end deep learning approaches. The key idea is to encapsulate the driving policy such that it is not directly exposed to raw perceptual input or low-level vehicle dynamics. We evaluate the presented approach in simulated urban environments and in the real world. In particular, we transfer a driving policy trained in simulation to a 1/5-scale robotic truck that is deployed in a variety of conditions, with no finetuning, on two continents. The supplementary video can be viewed at https://youtu.be/BrMDJqI6H5U

cs.RO↗

arXiv2018

Teaching UAVs to Race: End-to-End Regression of Agile Controls in Simulation

Automating the navigation of unmanned aerial vehicles (UAVs) in diverse scenarios has gained much attention in recent years. However, teaching UAVs to fly in challenging environments remains an unsolved problem, mainly due to the lack of training data. In this paper, we train a deep neural network to predict UAV controls from raw image data for the task of autonomous UAV racing in a photo-realistic simulation. Training is done through imitation learning with data augmentation to allow for the correction of navigation mistakes. Extensive experiments demonstrate that our trained network (when sufficient data augmentation is used) outperforms state-of-the-art methods and flies more consistently than many human pilots. Additionally, we show that our optimized network architecture can run in real-time on embedded hardware, allowing for efficient on-board processing critical for real-world deployment. From a broader perspective, our results underline the importance of extensive data augmentation techniques to improve robustness in end-to-end learning setups.

cs.CV↗

arXiv2018

Accurate experimental determination of Gallium K- and L3-shell XRF fundamental parameters

The fluorescence yield of the K- and L3-shell of gallium was determined using the radiometrically calibrated (reference-free) X-ray fluorescence instrumentation at the BESSY II synchrotron radiation facility. Simultaneous transmission and fluorescence signals from GaSe foils were obtained, resulting in K- and L3-shell fluorescence yield values consistent with existing database values(omega_Ga_K=0.515 +- 0.019, omega_Ga_L3=0.013 +- 0.001). For the first time, these standard combined uncertainties are obtained from a properly constructed Uncertainty Budget. These K-shell fluorescence yield values support Bambynek's semi-empirical compilation from 1972: these and other measurements yield a combined recommended value of omega_Ga_K=0.514 +- 0.010. Using the measured fluorescence yields together with production yields from reference Ga-implanted samples where the quantity of implanted Ga was determined at 1.3% traceable accuracy by Rutherford backscattering spectrometry, the K-shell and L3-subshell photoionization cross sections at selected incident photon energies were also determined and compared critically with the standard databases.

physics.atom-ph↗

arXiv2018

TrackingNet: A Large-Scale Dataset and Benchmark for Object Tracking in the Wild

Despite the numerous developments in object tracking, further development of current tracking algorithms is limited by small and mostly saturated datasets. As a matter of fact, data-hungry trackers based on deep-learning currently rely on object detection datasets due to the scarcity of dedicated large-scale tracking datasets. In this work, we present TrackingNet, the first large-scale dataset and benchmark for object tracking in the wild. We provide more than 30K videos with more than 14 million dense bounding box annotations. Our dataset covers a wide selection of object classes in broad and diverse context. By releasing such a large-scale dataset, we expect deep trackers to further improve and generalize. In addition, we introduce a new benchmark composed of 500 novel videos, modeled with a distribution similar to our training dataset. By sequestering the annotation of the test set and providing an online evaluation server, we provide a fair benchmark for future development of object trackers. Deep trackers fine-tuned on a fraction of our dataset improve their performance by up to 1.6% on OTB100 and up to 1.7% on TrackingNet Test. We provide an extensive benchmark on TrackingNet by evaluating more than 20 trackers. Our results suggest that object tracking in the wild is far from being solved.

cs.CV↗

arXiv2018

Sim4CV: A Photo-Realistic Simulator for Computer Vision Applications

We present a photo-realistic training and evaluation simulator (Sim4CV) with extensive applications across various fields of computer vision. Built on top of the Unreal Engine, the simulator integrates full featured physics based cars, unmanned aerial vehicles (UAVs), and animated human actors in diverse urban and suburban 3D environments. We demonstrate the versatility of the simulator with two case studies: autonomous UAV-based tracking of moving objects and autonomous driving using supervised learning. The simulator fully integrates both several state-of-the-art tracking algorithms with a benchmark evaluation tool and a deep neural network (DNN) architecture for training vehicles to drive autonomously. It generates synthetic photo-realistic datasets with automatic ground truth annotations to easily extend existing real-world datasets and provides extensive synthetic data variety through its ability to reconfigure synthetic worlds on the fly using an automatic world generation tool. The supplementary video can be viewed a https://youtu.be/SqAxzsQ7qUU

cs.CV↗

arXiv2018

End-to-end Driving via Conditional Imitation Learning

Deep networks trained on demonstrations of human driving have learned to follow roads and avoid obstacles. However, driving policies trained via imitation learning cannot be controlled at test time. A vehicle trained end-to-end to imitate an expert cannot be guided to take a specific turn at an upcoming intersection. This limits the utility of such systems. We propose to condition imitation learning on high-level command input. At test time, the learned driving policy functions as a chauffeur that handles sensorimotor coordination but continues to respond to navigational commands. We evaluate different architectures for conditional imitation learning in vision-based driving. We conduct experiments in realistic three-dimensional simulations of urban driving and on a 1/5 scale robotic truck that is trained to drive in a residential area. Both systems drive based on visual input yet remain responsive to high-level navigational commands. The supplementary video can be viewed at https://youtu.be/cFtnflNe5fM

cs.RO↗

arXiv2014

Wind Turbine Model and Observer in Takagi-Sugeno Model Structure

Based on a reduced-order, dynamic nonlinear wind turbine model in Takagi-Sugeno (TS) model structure, a TS state observer is designed as a disturbance observer to estimate the unknown effective wind speed. The TS observer model is an exact representation of the underlying nonlinear model, obtained by means of the sector-nonlinearity approach. The observer gain matrices are obtained by means of a linear matrix inequality (LMI) design approach for optimal fuzzy control, where weighting matrices for the individual system states and outputs are included. The observer is tested in simulations with the aero-elastic code FAST for the NREL 5 MW reference turbine, where it shows a stable behaviour both for IEC wind gusts and turbulent wind input.

eess.SY↗