搜索结果：Mayor

We present a comparative overview of state-of-the-art methods for modelling the distribution of neutral hydrogen (HI) in the post-reionization Universe, developed in preparation for upcoming SKAO cosmological surveys. Our aim is to assess how different physical and empirical assumptions reflect into predictions for key observables such as the cosmic HI density, the HI mass function, and the HI-halo mass relation. We consider both: (i) semi-analytical approaches that self-consistently evolve baryonic components within dark matter merger trees through physically motivated prescriptions and (ii) empirical schemes tailored to different observables and based on fast approximations designed for large ensemble studies. By comparing the predictions from the different methods considered, we find overall consistency in integrated quantities such as $Ω_{\mathrm{HI}}$, yet systematic differences in the detailed shape and scatter of the HI--halo mass relation and its redshift evolution. Semi-analytical models offer physically grounded predictions but depend on assumed prescriptions, while empirical methods provide flexibility and computational efficiency at the expense of robustness in extrapol

Upcoming galaxy surveys with the SKA Observatory will detect neutral hydrogen (HI) across unprecedented volumes, and their scientific return will crucially depend on predictive models for HI observables. In this work, we present a framework to simulate the neutral hydrogen 21cm emission line in such large-scale HI galaxy surveys. This framework is developed as a modular layer that builds on semi-analytical models. In particular we use as bases the Galaxy Evolution and Assembly (GAEA) and L-Galaxies semi-analytical models, coupled to merger trees from the Millennium Simulation suite. We validate our framework against local Universe observations, demonstrating consistency with velocity functions, and generalised Tully-Fisher relations. Predictions based on GAEA and L-Galaxies exhibit mutual consistency despite the distinct underlying physical prescriptions. We construct mock galaxy catalogues that incorporate forward-modelled selection functions, inclination effects, and redshift broadening, reproducing the statistical distributions of HI-selected galaxies in the ALFALFA survey. Finally, we present redshift distribution forecasts for future SKA Observatory HI galaxy surveys. This fra

The SKA Observatory will enable measurements of the Tully-Fisher relation for statistical samples of HI selected galaxies out to unprecedented depths and redshifts thanks to its unique combined spatial and spectral sensitivity. This chapter explores the transformative potential of such surveys for cosmology, in particular in the field of peculiar velocity measurements. We briefly review the present observational landscape for Tully-Fisher HI galaxy surveys and existing peculiar velocity datasets, and compare them with predictions for SKAO Tully-Fisher HI galaxy surveys with AA* and AA4 configurations of the SKA-Mid array. We discuss the extended range of cosmology science cases covered and enabled by such surveys.

The 21cm line from neutral hydrogen is expected to be a ubiquitous (albeit faint) tracer of galaxies in the late Universe. With SKAO-MID, large wide-field surveys of several million HI-containing galaxies will become feasible, resulting in catalogues of sufficient size to measure large-scale structure observables such as baryon acoustic oscillations and redshift-space distortions. While optical galaxy surveys over comparable areas are generally deeper, radio surveys of this kind have a number of other advantages, such as broader sampling of the halo mass function and the possibility of measuring luminosity distances via the Tully-Fisher relation. In this chapter, we provide predictions for the galaxy number counts versus redshift that will be achievable with a wide-field HI galaxy survey on SKAO-MID, along with corresponding forecasts for cosmological observables. Given the substantial uncertainty in the HI mass function with redshift, we bracket our predictions using a handful of different modelling methods.

Recent works have proposed accelerating the wall-clock training time of actor-critic methods via the use of large-scale environment parallelization; unfortunately, these can sometimes still require large number of environment interactions to achieve a desired level of performance. Noting that well-structured representations can improve the generalization and sample efficiency of deep reinforcement learning (RL) agents, we propose the use of simplicial embeddings: lightweight representation layers that constrain embeddings to simplicial structures. This geometric inductive bias results in sparse and discrete features that stabilize critic bootstrapping and strengthen policy gradients. When applied to FastTD3, FastSAC, and PPO, simplicial embeddings consistently improve sample efficiency and final performance across a variety of continuous- and discrete-control environments, without any loss in runtime speed.

Local administration of thrombolytics in ischemic stroke could accelerate clot lysis and the ensuing reperfusion while minimizing the side effects of systemic administration. Medical microrobots could be injected into the bloodstream and magnetically navigated to the clot for administering the drugs directly to the target. The magnetic manipulation required to navigate medical microrobots will depend on various parameters such as the microrobots size, the blood velocity, and the imposed magnetic field gradients. Numerical simulation was used to study the motion of magnetically controlled microrobots flowing through representative cerebral bifurcations, for predicting the magnetic gradients required to navigate the microrobots from the injection point until the target location. Upon thorough validation of the model against several independent analytical and experimental results, the model was used to generate maps and a predictive equation providing quantitative information on the required magnetic gradients, for different scenarios. The developed maps and predictive equation are crucial to inform the design, operation and optimization of magnetic navigation systems for healthcare a

The use of parallel actors for data collection has been an effective technique used in reinforcement learning (RL) algorithms. The manner in which data is collected in these algorithms, controlled via the number of parallel environments and the rollout length, induces a form of bias-variance trade-off; the number of training passes over the collected data, on the other hand, must strike a balance between sample efficiency and overfitting. We conduct an empirical analysis of these trade-offs on PPO, one of the most popular RL algorithms that uses parallel actors, and establish connections to network plasticity and, more generally, optimization stability. We examine its impact on network architectures, as well as the hyper-parameter sensitivity when scaling data. Our analyses indicate that larger dataset sizes can increase final performance across a variety of settings, and that scaling parallel environments is more effective than increasing rollout lengths. These findings highlight the critical role of data collection strategies in improving agent performance.

Significant research work has been undertaken related to the game-based learning approach over the last years. However, a closer look at this work reveals that further research is needed to examine some types of game-based learning approaches such as virtual reality serious games and LEGO Serious Play. This article examines and compares the effectiveness for learning Scrum and related agile practices of a serious game based on virtual reality and a learning activity based on the LEGO Serious Play methodology. The presented study used a quasi-experimental design with two groups, pre- and post-tests, and a perceptions questionnaire. The sample was composed of 59 software engineering students, 22 of which belonged to group A, while the other 37 were part of group B. The students in group A played the virtual reality serious game, whereas the students in group B conducted the LEGO Serious Play activity. The results show that both game-based learning approaches were effective for learning Scrum and related agile practices in terms of learning performance and motivation, but they also show that the students who played the virtual reality serious game outperformed their peers from the oth

In the last decades, important progress has been achieved in the understanding of the neurotrophic effects of intermittent fasting (IF), caloric restriction (CR) and exercise. Improved neuroprotection, synaptic plasticity and adult neurogenesis (NSPAN) are essential examples of these neurotrophic effects. The importance in this respect of the metabolic switch from glucose to ketone bodies as cellular fuel has been highlighted. More recently, calorie restriction mimetics (CRMs; resveratrol and other polyphenols in particular) have been investigated thoroughly in relation to NSPAN. In the narrative review sections of this manuscript, recent findings on these essential functions are synthesized and the most important molecules involved are presented. The most researched signaling pathways (PI3K, Akt, mTOR, AMPK, GSK3$β$, ULK, MAPK, PGC-1$α$, NF-$κ$B, sirtuins, Notch, Sonic hedgehog and Wnt) and processes (e.g., anti-inflammation, autophagy, apoptosis) that support or thwart neuroprotection, synaptic plasticity and neurogenesis are then briefly presented. This provides an accessible entry point to the literature. In the annotated bibliography section of this contribution, brief summari

Robust classification is essential in tasks like autonomous vehicle sign recognition, where the downsides of misclassification can be grave. Adversarial attacks threaten the robustness of neural network classifiers, causing them to consistently and confidently misidentify road signs. One such class of attack, shadow-based attacks, causes misidentifications by applying a natural-looking shadow to input images, resulting in road signs that appear natural to a human observer but confusing for these classifiers. Current defenses against such attacks use a simple adversarial training procedure to achieve a rather low 25\% and 40\% robustness on the GTSRB and LISA test sets, respectively. In this paper, we propose a robust, fast, and generalizable method, designed to defend against shadow attacks in the context of road sign recognition, that augments source images with binary adaptive threshold and edge maps. We empirically show its robustness against shadow attacks, and reformulate the problem to show its similarity to $\varepsilon$ perturbation-based attacks. Experimental results show that our edge defense results in 78\% robustness while maintaining 98\% benign test accuracy on the GT

We present some results of a CFHT adaptive optics search for companions to nearby dwarfs. We identify 21 new components in solar neighbourhood systems, of which 13 were found while surveying a volume-limited sample of M dwarfs within 12pc. We are obtaining complete observations for this subsample, to derive unbiased multiplicity statistics for the very-low-mass disk population. Additionally, we resolve for the first time 6 known spectroscopic or astrometric binaries, for a total of 27 newly resolved companions. A fair fraction of the new binaries has favourable parameters for accurate mass determinations. The newly resolved companion of Gl120.1C had an apparent spectroscopic minimum mass in the brown-dwarf range (Duquennoy & Mayor 1991) and it contributed to the statistical evidence that a few percent of solar type stars might have close-in brown-dwarf companions. We find that Gl~120.1C actually is an unrecognised double-lined spectroscopic pair. Its radial-velocity amplitude had therefore been strongly underestimated by Duquennoy & Mayor, and it does not truly belong to their sample of single-lined systems with minimum spectroscopic mass below the substellar limit. We also

Applying changes to an input speech signal to change the perceived speaker of speech to a target while maintaining the content of the input is a challenging but interesting task known as Voice conversion (VC). Over the last few years, this task has gained significant interest where most systems use data-driven machine learning models. Doing the conversion in a low-latency real-world scenario is even more challenging constrained by the availability of high-quality data. Data augmentations such as pitch shifting and noise addition are often used to increase the amount of data used for training machine learning based models for this task. In this paper we explore the efficacy of common data augmentation techniques for real-time voice conversion and introduce novel techniques for data augmentation based on audio and voice transformation effects as well. We evaluate the conversions for both male and female target speakers using objective and subjective evaluation methodologies.

Single molecule detection has revolutionised the fields of chemistry and biology by offering powerful ways to study individual molecules under different scenarios. Nanophotonic structures, including plasmonic antennas, significantly overcome the concentration limit at which single molecule events can be observed, enabling their detection at concentrations that are relevant to biological and chemical processes. Although antennas can be fabricated in large arrays, probing dynamic events requires high temporal resolution, which is best achieved by serial antenna interrogation. Unfortunately, this precludes the simultaneous recording from multiple antennas at different sample locations, and is time consuming, resulting in poor statistics and low-throughput data acquisition, abating the true potential of arrays. Here we exploit arrays of antenna-in-box nanostructures in combination with sCMOS readout to interrogate nanoscale volumes from 225 antennas simultaneously. Recording at 1 kHz allowed multiplexed dynamic measurements from 50 nanoantennas simultaneously with a temporal resolution dictated by the camera framerate and the photons emitted per molecule during a single passage. We dem

The stellar dynamics of Omega Centauri are inferred from the radial velocities of 469 stars measured with CORAVEL (Mayor et al. 1997). Rather than fit the data to a family of models, we generate estimates of all dynamical functions nonparametrically, by direct operation on the data. The cluster is assumed to be oblate and edge-on but mass is not assumed to follow light. The mean motions are consistent with axisymmetry but the rotation is not cylindrical. The peak rotational velocity is 7.9 km/s at 11 pc from the center. The apparent rotation of Omega Centauri is attributable in part to its proper motion. We reconstruct the stellar velocity ellipsoid as a function of position, assuming isotropy in the meridional plane. We find no significant evidence for a difference between the velocity dispersions parallel and perpendicular to the meridional plane. The mass distribution inferred from the kinematics is slightly more extended than, though not strongly inconsistent with, the luminosity distribution. We also derive the two-integral distribution function f(E,Lz) implied by the velocity data.

Since the discovery of the first giant planet outside the solar system in 1995 (Mayor & Queloz 1995), more than 180 extrasolar planets have been discovered. With improving detection capabilities, a new class of planets with masses 5-20 times larger than the Earth, at close distance from their parent star is rapidly emerging. Recently, the first system of three Neptune-mass planets has been discovered around the solar type star HD69830 (Lovis et al. 2006). Here, we present and discuss a possible formation scenario for this planetary system based on a consistent coupling between the extended core accretion model and evolutionary models (Alibert et al. 2005a, Baraffe et al. 2004,2006). We show that the innermost planet formed from an embryo having started inside the iceline is composed essentially of a rocky core surrounded by a tiny gaseous envelope. The two outermost planets started their formation beyond the iceline and, as a consequence, accrete a substantial amount of water ice during their formation. We calculate the present day thermodynamical conditions inside these two latter planets and show that they are made of a rocky core surrounded by a shell of fluid water and a ga

JWST has captured unusually detailed images of gas feeding the supermassive black hole at the center of NGC 4696。 A vast filament appears to funnel material into an 800-light-year-wide spinning disk, where gas races around at up to 600 kilometers per second。 The findings suggest black holes may recycle their own fuel by heating gas with jets and la

Grasshopper-like tests could begin in 2028