-
Unlocking the Potential of Operations Research for Multi-Graph Matching
Authors:
Max Kahl,
Sebastian Stricker,
Lisa Hutschenreiter,
Florian Bernard,
Bogdan Savchynskyy
Abstract:
We consider the incomplete multi-graph matching problem, which is a generalization of the NP-hard quadratic assignment problem for matching multiple finite sets. Multi-graph matching plays a central role in computer vision, e.g., for matching images or shapes, so that a number of dedicated optimization techniques have been proposed. While the closely related NP-hard multi-dimensional assignment pr…
▽ More
We consider the incomplete multi-graph matching problem, which is a generalization of the NP-hard quadratic assignment problem for matching multiple finite sets. Multi-graph matching plays a central role in computer vision, e.g., for matching images or shapes, so that a number of dedicated optimization techniques have been proposed. While the closely related NP-hard multi-dimensional assignment problem (MDAP) has been studied for decades in the operations research community, it only considers complete matchings and has a different cost structure. We bridge this gap and transfer well-known approximation algorithms for the MDAP to incomplete multi-graph matching. To this end, we revisit respective algorithms, adapt them to incomplete multi-graph matching, and propose their extended and parallelized versions. Our experimental validation shows that our new method substantially outperforms the previous state of the art in terms of objective and runtime. Our algorithm matches, for example, 29 images with more than 500 keypoints each in less than two minutes, whereas the fastest considered competitor requires at least half an hour while producing far worse results.
△ Less
Submitted 26 June, 2024;
originally announced June 2024.
-
Representation Learning for Appliance Recognition: A Comparison to Classical Machine Learning
Authors:
Matthias Kahl,
Daniel Jorde,
Hans-Arno Jacobsen
Abstract:
Non-intrusive load monitoring (NILM) aims at energy consumption and appliance state information retrieval from aggregated consumption measurements, with the help of signal processing and machine learning algorithms. Representation learning with deep neural networks is successfully applied to several related disciplines. The main advantage of representation learning lies in replacing an expert-driv…
▽ More
Non-intrusive load monitoring (NILM) aims at energy consumption and appliance state information retrieval from aggregated consumption measurements, with the help of signal processing and machine learning algorithms. Representation learning with deep neural networks is successfully applied to several related disciplines. The main advantage of representation learning lies in replacing an expert-driven, hand-crafted feature extraction with hierarchical learning from many representations in raw data format. In this paper, we show how the NILM processing-chain can be improved, reduced in complexity and alternatively designed with recent deep learning algorithms. On the basis of an event-based appliance recognition approach, we evaluate seven different classification models: a classical machine learning approach that is based on a hand-crafted feature extraction, three different deep neural network architectures for automated feature extraction on raw waveform data, as well as three baseline approaches for raw data processing. We evaluate all approaches on two large-scale energy consumption datasets with more than 50,000 events of 44 appliances. We show that with the use of deep learning, we are able to reach and surpass the performance of the state-of-the-art classical machine learning approach for appliance recognition with an F-Score of 0.75 and 0.86 compared to 0.69 and 0.87 of the classical approach.
△ Less
Submitted 26 August, 2022;
originally announced September 2022.
-
So2Sat POP -- A Curated Benchmark Data Set for Population Estimation from Space on a Continental Scale
Authors:
Sugandha Doda,
Yuanyuan Wang,
Matthias Kahl,
Eike Jens Hoffmann,
Kim Ouan,
Hannes Taubenböck,
Xiao Xiang Zhu
Abstract:
Obtaining a dynamic population distribution is key to many decision-making processes such as urban planning, disaster management and most importantly hel** the government to better allocate socio-technical supply. For the aspiration of these objectives, good population data is essential. The traditional method of collecting population data through the census is expensive and tedious. In recent y…
▽ More
Obtaining a dynamic population distribution is key to many decision-making processes such as urban planning, disaster management and most importantly hel** the government to better allocate socio-technical supply. For the aspiration of these objectives, good population data is essential. The traditional method of collecting population data through the census is expensive and tedious. In recent years, statistical and machine learning methods have been developed to estimate population distribution. Most of the methods use data sets that are either developed on a small scale or not publicly available yet. Thus, the development and evaluation of new methods become challenging. We fill this gap by providing a comprehensive data set for population estimation in 98 European cities. The data set comprises a digital elevation model, local climate zone, land use proportions, nighttime lights in combination with multi-spectral Sentinel-2 imagery, and data from the Open Street Map initiative. We anticipate that it would be a valuable addition to the research community for the development of sophisticated approaches in the field of population estimation.
△ Less
Submitted 10 November, 2022; v1 submitted 7 April, 2022;
originally announced April 2022.
-
Generative modeling of spatio-temporal weather patterns with extreme event conditioning
Authors:
Konstantin Klemmer,
Sudipan Saha,
Matthias Kahl,
Tianlin Xu,
Xiao Xiang Zhu
Abstract:
Deep generative models are increasingly used to gain insights in the geospatial data domain, e.g., for climate data. However, most existing approaches work with temporal snapshots or assume 1D time-series; few are able to capture spatio-temporal processes simultaneously. Beyond this, Earth-systems data often exhibit highly irregular and complex patterns, for example caused by extreme weather event…
▽ More
Deep generative models are increasingly used to gain insights in the geospatial data domain, e.g., for climate data. However, most existing approaches work with temporal snapshots or assume 1D time-series; few are able to capture spatio-temporal processes simultaneously. Beyond this, Earth-systems data often exhibit highly irregular and complex patterns, for example caused by extreme weather events. Because of climate change, these phenomena are only increasing in frequency. Here, we proposed a novel GAN-based approach for generating spatio-temporal weather patterns conditioned on detected extreme events. Our approach augments GAN generator and discriminator with an encoded extreme weather event segmentation mask. These segmentation masks can be created from raw input using existing event detection frameworks. As such, our approach is highly modular and can be combined with custom GAN architectures. We highlight the applicability of our proposed approach in experiments with real-world surface radiation and zonal wind data.
△ Less
Submitted 26 April, 2021;
originally announced April 2021.
-
Sine-Gordon on a wormhole
Authors:
Piotr Bizoń,
Maciej Dunajski,
Michał Kahl,
Michał Kowalczyk
Abstract:
In an attempt to understand the soliton resolution conjecture, we consider the Sine-Gordon equation on a spherically symmetric wormhole spacetime. We show that within each topological sector (indexed by a positive integer degree $n$) there exists a unique linearly stable soliton, which we call the $n$-kink. We give numerical evidence that the $n$-kink is a global attractor in the evolution of any…
▽ More
In an attempt to understand the soliton resolution conjecture, we consider the Sine-Gordon equation on a spherically symmetric wormhole spacetime. We show that within each topological sector (indexed by a positive integer degree $n$) there exists a unique linearly stable soliton, which we call the $n$-kink. We give numerical evidence that the $n$-kink is a global attractor in the evolution of any smooth, finite energy solutions of degree $n$. When the radius of the wormhole throat $a$ is large enough, the convergence to the $n$-kink is shown to be governed by internal modes that slowly decay due to the resonant transfer of energy to radiation. We compute the exact asymptotics of this relaxation process for the $1$-kink using the Soffer-Weinstein weakly nonlinear perturbation theory.
△ Less
Submitted 12 June, 2021; v1 submitted 21 December, 2020;
originally announced December 2020.
-
Training Auto-encoder-based Optimizers for Terahertz Image Reconstruction
Authors:
Tak Ming Wong,
Matthias Kahl,
Peter Haring Bolívar,
Andreas Kolb,
Michael Möller
Abstract:
Terahertz (THz) sensing is a promising imaging technology for a wide variety of different applications. Extracting the interpretable and physically meaningful parameters for such applications, however, requires solving an inverse problem in which a model function determined by these parameters needs to be fitted to the measured data. Since the underlying optimization problem is nonconvex and very…
▽ More
Terahertz (THz) sensing is a promising imaging technology for a wide variety of different applications. Extracting the interpretable and physically meaningful parameters for such applications, however, requires solving an inverse problem in which a model function determined by these parameters needs to be fitted to the measured data. Since the underlying optimization problem is nonconvex and very costly to solve, we propose learning the prediction of suitable parameters from the measured data directly. More precisely, we develop a model-based autoencoder in which the encoder network predicts suitable parameters and the decoder is fixed to a physically meaningful model function, such that we can train the encoding network in an unsupervised way. We illustrate numerically that the resulting network is more than 140 times faster than classical optimization techniques while making predictions with only slightly higher objective values. Using such predictions as starting points of local optimization techniques allows us to converge to better local minima about twice as fast as optimization without the network-based initialization.
△ Less
Submitted 29 October, 2019; v1 submitted 2 July, 2019;
originally announced July 2019.
-
Appliance Event Detection -- A Multivariate, Supervised Classification Approach
Authors:
Matthias Kahl,
Thomas Kriechbaumer,
Daniel Jorde,
Anwar Ul Haq,
Hans-Arno Jacobsen
Abstract:
Non-intrusive load monitoring (NILM) is a modern and still expanding technique, hel** to understand fundamental energy consumption patterns and appliance characteristics. Appliance event detection is an elementary step in the NILM pipeline. Unfortunately, several types of appliances (e.g., switching mode power supply (SMPS) or multi-state) are known to challenge state-of-the-art event detection…
▽ More
Non-intrusive load monitoring (NILM) is a modern and still expanding technique, hel** to understand fundamental energy consumption patterns and appliance characteristics. Appliance event detection is an elementary step in the NILM pipeline. Unfortunately, several types of appliances (e.g., switching mode power supply (SMPS) or multi-state) are known to challenge state-of-the-art event detection systems due to their noisy consumption profiles. Classical rule-based event detection system become infeasible and complex for these appliances. By step** away from distinct event definitions, we can learn from a consumer-configured event model to differentiate between relevant and irrelevant event transients.
We introduce a boosting oriented adaptive training, that uses false positives from the initial training area to reduce the number of false positives on the test area substantially. The results show a false positive decrease by more than a factor of eight on a dataset that has a strong focus on SMPS-driven appliances. To obtain a stable event detection system, we applied several experiments on different parameters to measure its performance. These experiments include the evaluation of six event features from the spectral and time domain, different types of feature space normalization to eliminate undesired feature weighting, the conventional and adaptive training, and two common classifiers with its optimal parameter settings. The evaluations are performed on two publicly available energy datasets with high sampling rates: BLUED and BLOND-50.
△ Less
Submitted 24 April, 2019;
originally announced April 2019.
-
Computational Image Enhancement for Frequency Modulated Continuous Wave (FMCW) THz Image
Authors:
Tak Ming Wong,
Matthias Kahl,
Peter Haring Bolívar,
Andreas Kolb
Abstract:
In this paper, a novel method to enhance Frequency Modulated Continuous Wave (FMCW) THz imaging resolution beyond its diffraction limit is proposed. Our method comprises two stages. Firstly, we reconstruct the signal in depth-direction using a sinc-envelope, yielding a significant improvement in depth estimation and signal parameter extraction. The resulting high precision depth estimate is used t…
▽ More
In this paper, a novel method to enhance Frequency Modulated Continuous Wave (FMCW) THz imaging resolution beyond its diffraction limit is proposed. Our method comprises two stages. Firstly, we reconstruct the signal in depth-direction using a sinc-envelope, yielding a significant improvement in depth estimation and signal parameter extraction. The resulting high precision depth estimate is used to deduce an accurate reflection intensity THz image. This image is fed in the second stage of our method to a 2D blind deconvolution procedure, adopted to enhance the lateral THz image resolution beyond the diffraction limit. Experimental data acquired with a FMCW system operating at 577 GHz with a bandwidth of 126 GHz shows that the proposed method enhances the lateral resolution by a factor of 2.29 to 346.2um with respect to the diffraction limit. The depth accuracy is 91um. Interestingly, the lateral resolution enhancement achieved with this blind deconvolution concept leads to better results in comparison to conventional gaussian deconvolution. Experimental data on a PCB resolution target is presented, in order to quantify the resolution enhancement and to compare the performance with established image enhancement approaches. The presented technique allows exposure of the interwoven fibre reinforced embedded structures of the PCB test sample.
△ Less
Submitted 2 July, 2019; v1 submitted 15 February, 2018;
originally announced February 2018.
-
A Yang-Mills field on the extremal Reissner-Nordström black hole
Authors:
Piotr Bizoń,
Michał Kahl
Abstract:
We consider a spherically symmetric (magnetic) $SU(2)$ Yang-Mills field propagating on the exterior of the extremal Reissner-Nordström black hole. Taking advantage of the conformal symmetry, we reduce the problem to the study of the Yang-Mills equation in a geodesically complete spacetime with two asymptotically flat ends. We prove the existence of infinitely many static solutions (two of which ar…
▽ More
We consider a spherically symmetric (magnetic) $SU(2)$ Yang-Mills field propagating on the exterior of the extremal Reissner-Nordström black hole. Taking advantage of the conformal symmetry, we reduce the problem to the study of the Yang-Mills equation in a geodesically complete spacetime with two asymptotically flat ends. We prove the existence of infinitely many static solutions (two of which are found in closed form) and determine the spectrum of their linear perturbations and quasinormal modes. Finally, using the hyperboloidal approach to the initial value problem, we describe the process of relaxation to the static endstates of evolution for various initial data.
△ Less
Submitted 15 March, 2016;
originally announced March 2016.
-
THz lattice vibrations for active plasmonics with light: Ultrafast optical response in gold/telluride hybrid plasmonic crystals
Authors:
Lars E. Kreilkamp,
Ilya A. Akimov,
Vladimir I. Belotelov,
Boris A. Glavin,
Leonid Litvin,
Axel Rudzinski,
Michael Kahl,
Ralf Jede,
Maciej Wiater,
Tomasz Wojtowicz,
Grzegorz Karczewski,
Dmitri R. Yakovlev,
Manfred Bayer
Abstract:
Excitation of coherent optical phonons in solids provides a pathway for ultrafast modulation of light on a sub-ps timescale. Here, we report on efficient 3.6 THz modulation of light reflected from hybrid metal/semiconductor plasmonic crystals caused by lattice vibrations in a few nm thick layer of elemental tellurium. We observe that surface plasmon polaritons contribute significantly to photoindu…
▽ More
Excitation of coherent optical phonons in solids provides a pathway for ultrafast modulation of light on a sub-ps timescale. Here, we report on efficient 3.6 THz modulation of light reflected from hybrid metal/semiconductor plasmonic crystals caused by lattice vibrations in a few nm thick layer of elemental tellurium. We observe that surface plasmon polaritons contribute significantly to photoinduced formation of this thin layer at the interface between a telluride-based II-VI semiconductor, such as (Cd,Mg)Te or (Cd,Mn)Te, and a one-dimensional gold grating. The change in interface composition is monitored via the excitation and detection of coherent optical tellurium phonons of $A_1$ symmetry by femtosecond laser pulses in a pump-probe experiment. The patterning of a plasmonic grating onto the semiconductor enhances the transient signal which originates from the interface region. This allows monitoring the layer formation and observing the shift of the phonon frequency caused by confinement of the lattice vibrations in the nm-thick segregated layer. Efficient excitation and detection of coherent optical phonons by means of surface plasmon polaritons are evidenced by the dependence of the signal strength on polarization of pump and probe pulses and its spectral distribution.
△ Less
Submitted 2 November, 2015;
originally announced November 2015.
-
Wave maps on a wormhole
Authors:
Piotr Bizoń,
Michał Kahl
Abstract:
We consider equivariant wave maps from a wormhole spacetime into the three-sphere. This toy-model is designed for gaining insight into the dissipation-by-dispersion phenomena, in particular the soliton resolution conjecture. We first prove that for each topological degree of the map there exists a unique static solution (harmonic map) which is linearly stable. Then, using the hyperboloidal formula…
▽ More
We consider equivariant wave maps from a wormhole spacetime into the three-sphere. This toy-model is designed for gaining insight into the dissipation-by-dispersion phenomena, in particular the soliton resolution conjecture. We first prove that for each topological degree of the map there exists a unique static solution (harmonic map) which is linearly stable. Then, using the hyperboloidal formulation of the initial value problem, we give numerical evidence that every solution starting from smooth initial data of any topological degree evolves asymptotically to the harmonic map of the same degree. The late-time asymptotics of this relaxation process is described in detail.
△ Less
Submitted 17 December, 2014;
originally announced December 2014.
