arXiv subjects

David Schneider

Publications and source records attributed to David Schneider.

30 records · Page 2Linked to original sources

arXiv2022

A Comparative Analysis of Decision-Level Fusion for Multimodal Driver Behaviour Understanding

Visual recognition inside the vehicle cabin leads to safer driving and more intuitive human-vehicle interaction but such systems face substantial obstacles as they need to capture different granularities of driver behaviour while dealing with highly limited body visibility and changing illumination. Multimodal recognition mitigates a number of such issues: prediction outcomes of different sensors complement each other due to different modality-specific strengths and weaknesses. While several late fusion methods have been considered in previously published frameworks, they constantly feature different architecture backbones and building blocks making it very hard to isolate the role of the chosen late fusion strategy itself. This paper presents an empirical evaluation of different paradigms for decision-level late fusion in video-based driver observation. We compare seven different mechanisms for joining the results of single-modal classifiers which have been both popular, (e.g. score averaging) and not yet considered (e.g. rank-level fusion) in the context of driver observation evaluating them based on different criteria and benchmark settings. This is the first systematic study of strategies for fusing outcomes of multimodal predictors inside the vehicles, conducted with the goal to provide guidance for fusion scheme selection.

cs.CV↗

arXiv2022

Discovery of a double detonation thermonuclear supernova progenitor

We present the discovery of a new double detonation progenitor system consisting of a hot subdwarf B (sdB) binary with a white dwarf companion with an P=76.34179(2) min orbital period. Spectroscopic observations are consistent with an sdB star during helium core burning residing on the extreme horizontal branch. Chimera light curves are dominated by ellipsoidal deformation of the sdB star and a weak eclipse of the companion white dwarf. Combining spectroscopic and light curve fits we find a low mass sdB star, $M_{\rm sdB}=0.383\pm0.028$ M$_\odot$ with a massive white dwarf companion, $M_{\rm WD}=0.725\pm0.026$ M$_\odot$. From the eclipses we find a blackbody temperature for the white dwarf of 26,800 K resulting in a cooling age of $\approx$25 Myrs whereas our MESA model predicts an sdB age of $\approx$170 Myrs. We conclude that the sdB formed first through stable mass transfer followed by a common envelope which led to the formation of the white dwarf companion $\approx$25 Myrs ago. Using the MESA stellar evolutionary code we find that the sdB star will start mass transfer in $\approx$6 Myrs and in $\approx$60 Myrs the white dwarf will reach a total mass of $0.92$ M$_\odot$ with a thick helium layer of $0.17$ M$_\odot$. This will lead to a detonation that will likely destroy the white dwarf in a peculiar thermonuclear supernova. PTF1 2238+7430 is only the second confirmed candidate for a double detonation thermonuclear supernova. Using both systems we estimate that at least $\approx$1% of white dwarf thermonuclear supernovae originate from sdB+WD binaries with thick helium layers, consistent with the small number of observed peculiar thermonuclear explosions.

astro-ph.SR↗

arXiv2021

Affect-DML: Context-Aware One-Shot Recognition of Human Affect using Deep Metric Learning

Human affect recognition is a well-established research area with numerous applications, e.g., in psychological care, but existing methods assume that all emotions-of-interest are given a priori as annotated training examples. However, the rising granularity and refinements of the human emotional spectrum through novel psychological theories and the increased consideration of emotions in context brings considerable pressure to data collection and labeling work. In this paper, we conceptualize one-shot recognition of emotions in context -- a new problem aimed at recognizing human affect states in finer particle level from a single support sample. To address this challenging task, we follow the deep metric learning paradigm and introduce a multi-modal emotion embedding approach which minimizes the distance of the same-emotion embeddings by leveraging complementary information of human appearance and the semantic scene context obtained through a semantic segmentation network. All streams of our context-aware model are optimized jointly using weighted triplet loss and weighted cross entropy loss. We conduct thorough experiments on both, categorical and numerical emotion recognition tasks of the Emotic dataset adapted to our one-shot recognition problem, revealing that categorizing human affect from a single example is a hard task. Still, all variants of our model clearly outperform the random baseline, while leveraging the semantic scene context consistently improves the learnt representations, setting state-of-the-art results in one-shot emotion recognition. To foster research of more universal representations of human affect states, we will make our benchmark and models publicly available to the community under https://github.com/KPeng9510/Affect-DML.

cs.CV↗

arXiv2021

preCICE v2: A Sustainable and User-Friendly Coupling Library

preCICE is a free/open-source coupling library. It enables creating partitioned multi-physics simulations by gluing together separate software packages. This paper summarizes the development efforts in preCICE of the past five years. During this time span, we have turned the software from a working prototype -- sophisticated numerical coupling methods and scalability on ten thousands of compute cores -- to a sustainable and user-friendly software project with a steadily-growing community. Today, we know through forum discussions, conferences, workshops, and publications of more than 100 research groups using preCICE. We cover the fundamentals of the software alongside a performance and accuracy analysis of different data mapping methods. Afterwards, we describe ready-to-use integration with widely-used external simulation software packages, tests and continuous integration from unit to system level, and community building measures, drawing an overview of the current preCICE ecosystem.

cs.MS↗

arXiv2021

Let's Play for Action: Recognizing Activities of Daily Living by Learning from Life Simulation Video Games

Recognizing Activities of Daily Living (ADL) is a vital process for intelligent assistive robots, but collecting large annotated datasets requires time-consuming temporal labeling and raises privacy concerns, e.g., if the data is collected in a real household. In this work, we explore the concept of constructing training examples for ADL recognition by playing life simulation video games and introduce the SIMS4ACTION dataset created with the popular commercial game THE SIMS 4. We build Sims4Action by specifically executing actions-of-interest in a "top-down" manner, while the gaming circumstances allow us to freely switch between environments, camera angles and subject appearances. While ADL recognition on gaming data is interesting from the theoretical perspective, the key challenge arises from transferring it to the real-world applications, such as smart-homes or assistive robotics. To meet this requirement, Sims4Action is accompanied with a GamingToReal benchmark, where the models are evaluated on real videos derived from an existing ADL dataset. We integrate two modern algorithms for video-based activity recognition in our framework, revealing the value of life simulation video games as an inexpensive and far less intrusive source of training data. However, our results also indicate that tasks involving a mixture of gaming and real data are challenging, opening a new research direction. We will make our dataset publicly available at https://github.com/aroitberg/sims4action.

cs.CV↗

arXiv2020

Runaway blue main-sequence stars at high Galactic latitudes. Target selection with Gaia and spectroscopic identification

Motivated by the historical identification of runaway main-sequence (MS) stars of early spectral type at high Galactic latitudes, we test the capability of Gaia at identifying new such stars. We have selected ~2300 sources with Gaia magnitudes of GBP - GRP < 0.05, compatible with the colors of low-extinction MS stars earlier than mid-A spectral type, and obtained low-resolution optical spectroscopy for 48 such stars. By performing detailed photometric and spectroscopic analyses, we derive their atmospheric and physical parameters (effective temperature, surface gravity, radial velocity, interstellar reddening, spectrophotometric distance, mass, radius, luminosity, and age). The comparison between spectrophotometric and parallax-based distances enables us to disentangle the MS candidates from older blue horizontal branch (BHB) candidates. We identify 12 runaway MS candidates, with masses between 2 and 6 Msun. Their trajectories are traced back to the Galactic disc to identify their most recent Galactic plane crossings and the corresponding flight times. All 12 candidates are ejected from the Galactic disc within 2 to 16.5 kpc from the Galactic center and possess flight times that are shorter than their evolutionary ages, compatible with a runaway hypothesis. Three MS candidates have ejection velocities exceeding 450 km/s, thus, appear to challenge the canonical ejection scenarios for late B-type stars. The fastest star of our sample also has a non-negligible Galactic escape probability if its MS nature can be confirmed. We identify 27 BHB candidates, and the two hottest stars in our sample are rare late O and early B type stars of low mass evolving towards the white dwarf cooling sequence.

astro-ph.SR↗

arXiv2020

EVR-CB-004: An Inflated Hot Subdwarf O star + Unseen WD Companion in a Compact Binary Discovered with the Evryscope

We present the discovery of EVR-CB-004, a close binary with a remnant stellar core and an unseen white dwarf companion. The analysis in this work reveals the primary is potentially an inflated hot subdwarf (sdO) and more likely is a rarer post-blue horizontal branch (post-BHB) star. Post-BHBs are the short-lived shell-burning final stage of a blue horizontal star or hot subdwarf before transitioning to a WD. This object was discovered using Evryscope photometric data in a southern-all-sky hot subdwarf variability survey. The photometric light curve for EVR-CB-004 shows multi-component variability from ellipsoidal deformation of the primary and from Doppler boosting as well as gravitational limb darkening. EVR-CB-004 is one of just a handful of known systems, and has a long period (6.08426 hours) and large amplitude ellipsoidal modulation (16.0 $\%$ change in brightness from maximum to minimum) for these extremely close binary systems, while the properties of the primary make it a truly unique system. EVR-CB-004 also shows a peculiar low-amplitude (less than $1\%$) sinusoidal light curve variation with a period that is a 1/3 resonance of the binary period. We tentatively identify this additional variation source as a tidally-induced resonant pulsation, and we suggest followup observations that could verify this interpretation. From the evolutionary state of the system, its components, and its mass fraction, EVR-CB-004 is a strong merger candidate to form a single high-mass ($\approx1.2M_{\odot}$) WD. EVR-CB-004 offers a glimpse into a brief phase of a remnant core evolution and secondary variation, not seen before in a compact binary.

astro-ph.SR↗

arXiv2020

On the Performance of Bytecode Interpreters in Prolog

The semantics and the recursive execution model of Prolog make it very natural to express language interpreters in form of AST (Abstract Syntax Tree) interpreters where the execution follows the tree representation of a program. An alternative implementation technique is that of bytecode interpreters. These interpreters transform the program into a compact and linear representation before evaluating it and are generally considered to be faster and to make better use of resources. In this paper, we discuss different ways to express the control flow of interpreters in Prolog and present several implementations of AST and bytecode interpreters. On a simple language designed for this purpose, we evaluate whether techniques best known from imperative languages are applicable in Prolog and how well they perform. Our ultimate goal is to assess which interpreter design in Prolog is the most efficient, as we intend to apply these results to a more complex language. However, we believe the analysis in this paper to be of more general interest.

cs.PL↗

arXiv2020

A new class of Roche lobe-filling hot subdwarf binaries

We present the discovery of the second binary with a Roche lobe-filling hot subdwarf transferring mass to a white dwarf (WD) companion. This 56 minute binary was discovered using data from the Zwicky Transient Facility. Spectroscopic observations reveal an He-sdOB star with an effective temperature of $T_{\rm eff}=33,700\pm1000$ K and a surface gravity of $log(g)=5.54\pm0.11$. The GTC+HiPERCAM light curve is dominated by the ellipsoidal deformation of the He-sdOB star and shows an eclipse of the He-sdOB by an accretion disk as well as a weak eclipse of the WD. We infer a He-sdOB mass of $M_{\rm sdOB}=0.41\pm0.04$ M$_\odot$ and a WD mass of $M_{\rm WD}=0.68\pm0.05$ M$_\odot$. The weak eclipses imply a WD black-body temperature of $63,000\pm10,000$ K and a radius $R_{\rm WD}=0.0148\pm0.0020$ M$_\odot$ as expected for a WD of such high temperature. The He-sdOB star is likely undergoing hydrogen shell burning and will continue transferring mass for $\approx1$ Myrs at a rate of $10^{-9} M_\odot {\rm yr}^{-1}$ which is consistent with the high WD temperature. The hot subdwarf will then turn into a WD and the system will merge in $\approx30$ Myrs. We suggest that Galactic reddening could bias discoveries towards preferentially finding Roche lobe-filling systems during the short-lived shell burning phase. Studies using reddening corrected samples should reveal a large population of helium core-burning hot subdwarfs with $T_{\rm eff}\approx25,000$ K in binaries of 60-90 minutes with WDs. Though not yet in contact, these binaries would eventually come into contact through gravitational wave emission and explode as a sub-luminous thermonuclear supernova or evolve into a massive single WD.

astro-ph.SR↗

arXiv2019

An Interactive Indoor Drone Assistant

With the rapid advance of sophisticated control algorithms, the capabilities of drones to stabilise, fly and manoeuvre autonomously have dramatically improved, enabling us to pay greater attention to entire missions and the interaction of a drone with humans and with its environment during the course of such a mission. In this paper, we present an indoor office drone assistant that is tasked to run errands and carry out simple tasks at our laboratory, while given instructions from and interacting with humans in the space. To accomplish its mission, the system has to be able to understand verbal instructions from humans, and perform subject to constraints from control and hardware limitations, uncertain localisation information, unpredictable and uncertain obstacles and environmental factors. We combine and evaluate the dialogue, navigation, flight control, depth perception and collision avoidance components. We discuss performance and limitations of our assistant at the component as well as the mission level. A 78% mission success rate was obtained over the course of 27 missions.

cs.RO↗

arXiv2014

Two-strain competition in quasi-neutral stochastic disease dynamics

We develop a new perturbation method for studying quasi-neutral competition in a broad class of stochastic competition models, and apply it to the analysis of fixation of competing strains in two epidemic models. The first model is a two-strain generalization of the stochastic Susceptible-Infected-Susceptible (SIS) model. Here we extend previous results due to Parsons and Quince (2007), Parsons et al (2008) and Lin, Kim and Doering (2012). The second model, a two-strain generalization of the stochastic Susceptible-Infected-Recovered (SIR) model with population turnover, has not been studied previously. In each of the two models, when the basic reproduction numbers of the two strains are identical, a system with an infinite population size approaches a point on the deterministic coexistence line (CL): a straight line of fixed points in the phase space of sub-population sizes. Shot noise drives one of the strain populations to fixation, and the other to extinction, on a time scale proportional to the total population size. Our perturbation method explicitly tracks the dynamics of the probability distribution of the sub-populations in the vicinity of the CL. We argue that, whereas the slow strain has a competitive advantage for mathematically "typical" initial conditions, it is the fast strain that is more likely to win in the important situation when a few infectives of both strains are introduced into a susceptible population.

cond-mat.stat-mech↗

arXiv2009

A Semantics-Aware Editing Environment for Prolog in Eclipse

In this paper we present a Prolog plugin for Eclipse based upon BE4, and providing many features such as semantic-aware syntax highlighting, outline view, error marking, content assist, hover information, documentation generation, and quick fixes. The plugin makes use of a Java parser for full Prolog with an integrated Prolog engine, and can be extended with further semantic analyses, e.g., based on abstract interpretation.

cs.PL↗