arXiv subjects

Christian Wolf

Publications and source records attributed to Christian Wolf.

At least 73 records · Page 4Linked to original sources

arXiv2022

Discovery of the most luminous quasar of the last 9 Gyr

We report the discovery of a bright (g = 14.5 mag (AB), K = 11.9 mag (Vega)) quasar at redshift z = 0.83 -- the optically brightest (unbeamed) quasar at z > 0.4. SMSS J114447.77-430859.3, at a Galactic latitude of b = +18.1deg, was identified by its optical colours from the SkyMapper Southern Survey (SMSS) during a search for symbiotic binary stars. Optical and near-infrared spectroscopy reveals broad MgII, H-beta, H-alpha, and Pa-beta emission lines, from which we measure a black hole mass of log10(M_BH/M_Sun) = 9.4 +/- 0.5. With its high luminosity, L_bol = (4.7 +/- 1.0) * 10^47 erg/s or M_i(z=2) = -29.74 mag (AB), we estimate an Eddington ratio of ~1.4. As the most luminous quasar known over the last ~9 Gyr of cosmic history, having a luminosity 8 times greater than 3C 273, the source offers a range of potential follow-up opportunities.

astro-ph.GA↗

arXiv2022

A Roche Lobe-filling hot Subdwarf and White Dwarf Binary: possible detection of an ejected common envelope

Binaries consisting of a hot subdwarf star and an accreting white dwarf (WD) are sources of gravitational wave radiation at low frequencies and possible progenitors of type Ia supernovae if the WD mass is large enough. Here, we report the discovery of the third binary known of this kind: it consists of a hot subdwarf O (sdO) star and a WD with an orbital period of 3.495 hours and an orbital shrinkage of 0.1 s in 6 yr. The sdO star overfills its Roche lobe and likely transfers mass to the WD via an accretion disk. From spectroscopy, we obtain an effective temperature of $T_{\mathrm{eff}}=54\,240\pm1\,840$ K and a surface gravity of $\log{g}=4.841\pm0.108$ for the sdO star. From the light curve analysis, we obtain a sdO mass of $M_{\mathrm{sdO}}=0.55$ ${\mathrm{M_{\odot}}}$ and a mass ratio of $q=M_{\mathrm{WD}}/M_{\mathrm{sdO}}=0.738\pm0.001$. Also, we estimate that the disk has a radius of $\sim 0.41R_\odot$ and a thickness of $\sim 0.18R_\odot$. The origin of this binary is probably a common envelope ejection channel, where the progenitor of the sdO star is either an RGB star or, more likely, an early AGB star; the sdO star will subsequently evolve into a WD and merge with its WD companion, likely resulting in an R CrB star. The outstanding feature in the spectrum of this object is strong Ca H&K lines, which are blueshifted by $\sim$200 km/s and likely originate from the recently ejected common envelope, and we estimated that the remnant CE material in the binary system has a density $\sim 6\times 10^{-10} {\rm g/cm^3}$.

astro-ph.SR↗

arXiv2022

Finding Quasars behind the Galactic Plane. II. Spectroscopic Identifications of 204 Quasars at $|b|< 20°$

Quasars behind the Galactic plane (GPQs) are important astrometric references and valuable probes of Galactic gas, yet the search for GPQs is difficult due to severe extinction and source crowding in the Galactic plane. In this paper, we present a sample of 204 spectroscopically confirmed GPQs at |b|<20°, 191 of which are new discoveries. This GPQ sample covers a wide redshift range from 0.069 to 4.487. For the subset of 230 observed GPQ candidates, the lower limit of the purity of quasars is 85.2%, and the lower limit of the fraction of stellar contaminants is 6.1%. Using a multicomponent spectral fitting, we measure the emission line and continuum flux of the GPQs, and estimate their single-epoch virial black hole masses. Due to selection effects raised from Galactic extinction and target magnitude, these GPQs have higher black hole masses and continuum luminosities in comparison to the SDSS DR7 quasar sample. The spectral-fitting results and black hole mass estimates are compiled into a main spectral catalog, and an extended spectral catalog of GPQs. The successful identifications prove the reliability of both our GPQ selection methods and the GPQ candidate catalog, shedding light on the astrometric and astrophysical programs that make use of a large sample of GPQs in the future.

astro-ph.GA↗

arXiv2022

Deep Extragalactic VIsible Legacy Survey (DEVILS): The emergence of bulges and decline of disk growth since $z = 1$

We present a complete structural analysis of the ellipticals (E), diffuse bulges (dB), compact bulges (cB), and disks (D) within a redshift range $0 < z < 1$, and stellar mass $\log_{10}(\mathrm{M}_*/\mathrm{M}_\odot) \geq 9.5$ volume-limited sample drawn from the combined DEVILS and HST-COSMOS region. We use the {\sc ProFit} code to profile over $\sim35,000$ galaxies for which visual classification into single or double-component was predefined in Paper-I. Over this redshift range, we see a growth in the total stellar mass density (SMD) of a factor of 1.5. At all epochs we find that the dominant structure, contributing to the total SMD, is the disk, and holds a fairly constant share of $\sim60\%$ of the total SMD from $z = 0.8$ to $z = 0.2$, dropping to $\sim30\%$ at $z = 0.0$ (representing $\sim33\%$ decline in the total disk SMD). Other classes (E, dB, and cB) show steady growth in their numbers and integrated stellar mass densities. By number, the most dramatic change across the full mass range is in the growth of diffuse bulges. In terms of total SMD, the biggest gain is an increase in massive elliptical systems, rising from 20\% at $z = 0.8$ to equal that of disks at $z = 0.0$ (30\%) representing an absolute mass growth of a factor of 2.5. Overall we see a clear picture of the emergence and growth of all three classes of spheroids over the past 8 Gyrs, and infer that in the later half of the Universe's timeline spheroid forming-processes and pathways (secular evolution, mass-accretion, and mergers) appear to dominate mass transformation over quiescent disk growth.

astro-ph.GA↗

arXiv2022

Learning Reduced Nonlinear State-Space Models: an Output-Error Based Canonical Approach

The identification of a nonlinear dynamic model is an open topic in control theory, especially from sparse input-output measurements. A fundamental challenge of this problem is that very few to zero prior knowledge is available on both the state and the nonlinear system model. To cope with this challenge, we investigate the effectiveness of deep learning in the modeling of dynamic systems with nonlinear behavior by advocating an approach which relies on three main ingredients: (i) we show that under some structural conditions on the to-be-identified model, the state can be expressed in function of a sequence of the past inputs and outputs; (ii) this relation which we call the state map can be modelled by resorting to the well-documented approximation power of deep neural networks; (iii) taking then advantage of existing learning schemes, a state-space model can be finally identified. After the formulation and analysis of the approach, we show its ability to identify three different nonlinear systems. The performances are evaluated in terms of open-loop prediction on test data generated in simulation as well as a real world data-set of unmanned aerial vehicle flight measurements.

eess.SY↗

arXiv2022

Chemical Abundance of z~6 Quasar Broad-Line Regions in the XQR-30 Sample

The elemental abundances in the broad-line regions of high-redshift quasars trace the chemical evolution in the nuclear regions of massive galaxies in the early universe. In this work, we study metallicity-sensitive broad emission-line flux ratios in rest-frame UV spectra of 25 high-redshift (5.8 < z < 7.5) quasars observed with the VLT/X-shooter and Gemini/GNIRS instruments, ranging over $\log(M_{\rm{BH}}/M_{\odot})= 8.4-9.8$ in black hole mass and $\log(L_{\rm{bol}}/\rm{erg\, s}^{-1})= 46.7-47.7$ in bolometric luminosity. We fit individual spectra and composites generated by binning across quasar properties: bolometric luminosity, black hole mass, and blueshift of the \civ\, line, finding no redshift evolution in the emission-line ratios by comparing our high-redshift quasars to lower-redshift (2.0 < z < 5.0) results presented in the literature. Using Cloudy-based locally optimally-emitting cloud photoionisation model relations between metallicity and emission-line flux ratios, we find the observable properties of the broad emission lines to be consistent with emission from gas clouds with metallicity that are at least 2-4 times solar. Our high-redshift measurements also confirm that the blueshift of the CIV emission line is correlated with its equivalent width, which influences line ratios normalised against CIV. When accounting for the CIV blueshift, we find that the rest-frame UV emission-line flux ratios do not correlate appreciably with the black hole mass or bolometric luminosity.

astro-ph.GA↗

arXiv2022

Learning to estimate UAV created turbulence from scene structure observed by onboard cameras

Controlling UAV flights precisely requires a realistic dynamic model and accurate state estimates from onboard sensors like UAV, GPS and visual observations. Obtaining a precise dynamic model is extremely difficult, as important aerodynamic effects are hard to model, in particular ground effect and other turbulences. While machine learning has been used in the past to estimate UAV created turbulence, this was restricted to flat grounds or diffuse in-flight air turbulences, both without taking into account obstacles. In this work we address the complex problem of estimating in-flight turbulences caused by obstacles, in particular the complex structures in cluttered environments. We learn a mapping from control input and images captured by onboard cameras to turbulence. In a large-scale setting, we train a model over a large number of different simulated photo-realistic environments loaded into the Habitat.AI simulator augmented with a dynamic UAV model and an analytic ground effect model. We transfer the model from simulation to a real environment and evaluate on real UAV flights from the EuRoC-MAV dataset, showing that the model is capable of good sim2real generalization performance. The dataset will be made publicly available upon acceptance.

cs.RO↗

arXiv2022

An experimental study of the vision-bottleneck in VQA

As in many tasks combining vision and language, both modalities play a crucial role in Visual Question Answering (VQA). To properly solve the task, a given model should both understand the content of the proposed image and the nature of the question. While the fusion between modalities, which is another obviously important part of the problem, has been highly studied, the vision part has received less attention in recent work. Current state-of-the-art methods for VQA mainly rely on off-the-shelf object detectors delivering a set of object bounding boxes and embeddings, which are then combined with question word embeddings through a reasoning module. In this paper, we propose an in-depth study of the vision-bottleneck in VQA, experimenting with both the quantity and quality of visual objects extracted from images. We also study the impact of two methods to incorporate the information about objects necessary for answering a question, in the reasoning module directly, and earlier in the object selection stage. This work highlights the importance of vision in the context of VQA, and the interest of tailoring vision methods used in VQA to the task at hand.

cs.CV↗

arXiv2022

SkyMapper colours of Seyfert galaxies and Changing-look AGN-II. Newly discovered Changing-look AGN

Changing-look Active Galactic Nuclei (CLAGN) are AGN that change type as their broad emission lines appear or disappear, which is usually accompanied by strong flux changes inn their blue featureless continuum. We search for Turn-On CLAGN by selecting type-2 AGN from the spectroscopic 6dF Galaxy Survey, whose photometry, as observed over a decade later by the SkyMapper Southern Survey, is consistent with type-1 AGN. Starting from a random sample of 235 known type-2 AGN we select 18 candidates and confirm 13 AGN to have changed into type-1 spectra; observations of an incomplete sample reveal nine further Turn-On CLAGN. While our search was not intended to reliably discover Turn-Off CLAGN, we discover four such cases as well. This result suggests a Turn-On CLAGN rate of 12% over ~15 years and imply a total CLAGN rate of ~25% over this period. Finally, we present observations of AGN that are atypical for the CLAGN phenomenology, including J1109146 - a CLAGN that did not appear as an AGN in 6dFGS; J1406507 - the second reported Changinglook NLS1; and J1340153 - a CLAGN with a change timescale of three months.

astro-ph.GA↗

arXiv2022

Filtered-CoPhy: Unsupervised Learning of Counterfactual Physics in Pixel Space

Learning causal relationships in high-dimensional data (images, videos) is a hard task, as they are often defined on low dimensional manifolds and must be extracted from complex signals dominated by appearance, lighting, textures and also spurious correlations in the data. We present a method for learning counterfactual reasoning of physical processes in pixel space, which requires the prediction of the impact of interventions on initial conditions. Going beyond the identification of structural relationships, we deal with the challenging problem of forecasting raw video over long horizons. Our method does not require the knowledge or supervision of any ground truth positions or other object or scene properties. Our model learns and acts on a suitable hybrid latent representation based on a combination of dense features, sets of 2D keypoints and an additional latent vector per keypoint. We show that this better captures the dynamics of physical processes than purely dense or sparse representations. We introduce a new challenging and carefully designed counterfactual benchmark for predictions in pixel space and outperform strong baselines in physics-inspired ML and video prediction.

cs.CV↗

arXiv2022

MoCap-less Quantitative Evaluation of Ego-Pose Estimation Without Ground Truth Measurements

The emergence of data-driven approaches for control and planning in robotics have highlighted the need for developing experimental robotic platforms for data collection. However, their implementation is often complex and expensive, in particular for flying and terrestrial robots where the precise estimation of the position requires motion capture devices (MoCap) or Lidar. In order to simplify the use of a robotic platform dedicated to research on a wide range of indoor and outdoor environments, we present a data validation tool for ego-pose estimation that does not require any equipment other than the on-board camera. The method and tool allow a rapid, visual and quantitative evaluation of the quality of ego-pose sensors and are sensitive to different sources of flaws in the acquisition chain, ranging from desynchronization of the sensor flows to misevaluation of the geometric parameters of the robotic platform. Using computer vision, the information from the sensors is used to calculate the motion of a semantic scene point through its projection to the 2D image space of the on-board camera. The deviations of these keypoints from references created with a semi-automatic tool allow rapid and simple quality assessment of the data collected on the platform. To demonstrate the performance of our method, we evaluate it on two challenging standard UAV datasets as well as one dataset taken from a terrestrial robot.

cs.RO↗

arXiv2022

Deep KKL: Data-driven Output Prediction for Non-Linear Systems

We address the problem of output prediction, ie. designing a model for autonomous nonlinear systems capable of forecasting their future observations. We first define a general framework bringing together the necessary properties for the development of such an output predictor. In particular, we look at this problem from two different viewpoints, control theory and data-driven techniques (machine learning), and try to formulate it in a consistent way, reducing the gap between the two fields. Building on this formulation and problem definition, we propose a predictor structure based on the Kazantzis-Kravaris/Luenberger (KKL) observer and we show that KKL fits well into our general framework. Finally, we propose a constructive solution for this predictor that solely relies on a small set of trajectories measured from the system. Our experiments show that our solution allows to obtain an efficient predictor over a subset of the observation space.

cs.LG↗

arXiv2021

Graph augmented Deep Reinforcement Learning in the GameRLand3D environment

We address planning and navigation in challenging 3D video games featuring maps with disconnected regions reachable by agents using special actions. In this setting, classical symbolic planners are not applicable or difficult to adapt. We introduce a hybrid technique combining a low level policy trained with reinforcement learning and a graph based high level classical planner. In addition to providing human-interpretable paths, the approach improves the generalization performance of an end-to-end approach in unseen maps, where it achieves a 20% absolute increase in success rate over a recurrent end-to-end agent on a point to point navigation task in yet unseen large-scale maps of size 1km x 1km. In an in-depth experimental study, we quantify the limitations of end-to-end Deep RL approaches in vast environments and we also introduce "GameRLand3D", a new benchmark and soon to be released environment can generate complex procedural 3D maps for navigation tasks.

cs.LG↗

arXiv2021

Ultra-luminous high-redshift quasars from SkyMapper -- II. New quasars and the bright end of the luminosity function

We search for ultra-luminous Quasi-Stellar Objects (QSOs) at high redshift using photometry from the SkyMapper Southern Survey Data Release 3 (DR3), in combination with 2MASS, VHS DR6, VIKING DR5, AllWISE, and CatWISE2020, as well as parallaxes and proper motions from Gaia DR2 and eDR3. We report 142 newly discovered Southern QSOs at $3.8 80$ per cent complete for $M_{145}<-27.33$ ABmag at $z=4.7$ and for $M_{145}<-27.73$ ABmag at $z=5.4$. We derive the bright end of the QSO luminosity function at restframe 145 nm for $z=4.7-5.4$ and measure its slope to be $β= -3.60\pm0.37$ and $β= -3.38\pm0.32$ for two different estimates of the faint-end QSO density adopted from the literature. We also present the first $z\sim 5$ QSO luminosity function at restframe 300 nm.

astro-ph.CO↗

arXiv2021

Godot Reinforcement Learning Agents

We present Godot Reinforcement Learning (RL) Agents, an open-source interface for developing environments and agents in the Godot Game Engine. The Godot RL Agents interface allows the design, creation and learning of agent behaviors in challenging 2D and 3D environments with various on-policy and off-policy Deep RL algorithms. We provide a standard Gym interface, with wrappers for learning in the Ray RLlib and Stable Baselines RL frameworks. This allows users access to over 20 state of the art on-policy, off-policy and multi-agent RL algorithms. The framework is a versatile tool that allows researchers and game designers the ability to create environments with discrete, continuous and mixed action spaces. The interface is relatively performant, with 12k interactions per second on a high end laptop computer, when parallized on 4 CPU cores. An overview video is available here: https://youtu.be/g1MlZSFqIj4

cs.LG↗

arXiv2021

SIM2REALVIZ: Visualizing the Sim2Real Gap in Robot Ego-Pose Estimation

The Robotics community has started to heavily rely on increasingly realistic 3D simulators for large-scale training of robots on massive amounts of data. But once robots are deployed in the real world, the simulation gap, as well as changes in the real world (e.g. lights, objects displacements) lead to errors. In this paper, we introduce Sim2RealViz, a visual analytics tool to assist experts in understanding and reducing this gap for robot ego-pose estimation tasks, i.e. the estimation of a robot's position using trained models. Sim2RealViz displays details of a given model and the performance of its instances in both simulation and real-world. Experts can identify environment differences that impact model predictions at a given location and explore through direct interactions with the model hypothesis to fix it. We detail the design of the tool, and case studies related to the exploit of the regression to the mean bias and how it can be addressed, and how models are perturbed by the vanish of landmarks such as bikes.

cs.RO↗

arXiv2021

An in-depth experimental study of sensor usage and visual reasoning of robots navigating in real environments

Visual navigation by mobile robots is classically tackled through SLAM plus optimal planning, and more recently through end-to-end training of policies implemented as deep networks. While the former are often limited to waypoint planning, but have proven their efficiency even on real physical environments, the latter solutions are most frequently employed in simulation, but have been shown to be able learn more complex visual reasoning, involving complex semantical regularities. Navigation by real robots in physical environments is still an open problem. End-to-end training approaches have been thoroughly tested in simulation only, with experiments involving real robots being restricted to rare performance evaluations in simplified laboratory conditions. In this work we present an in-depth study of the performance and reasoning capacities of real physical agents, trained in simulation and deployed to two different physical environments. Beyond benchmarking, we provide insights into the generalization capabilities of different agents training in different conditions. We visualize sensor usage and the importance of the different types of signals. We show, that for the PointGoal task, an agent pre-trained on wide variety of tasks and fine-tuned on a simulated version of the target environment can reach competitive performance without modelling any sim2real transfer, i.e. by deploying the trained agent directly from simulation to a real physical robot.

cs.AI↗

arXiv2021

Satellite Image Semantic Segmentation

In this paper, we propose a method for the automatic semantic segmentation of satellite images into six classes (sparse forest, dense forest, moor, herbaceous formation, building, and road). We rely on Swin Transformer architecture and build the dataset from IGN open data. We report quantitative and qualitative segmentation results on this dataset and discuss strengths and limitations. The dataset and the trained model are made publicly available.

cs.CV↗