Search

Monitoring air pollution is crucial for protecting human health from exposure to harmful substances. Traditional methods of air quality monitoring, such as ground-based sensors and satellite-based remote sensing, face limitations due to high deployment costs, sparse sensor coverage, and environmental interferences. To address these challenges, this paper proposes a framework for high-resolution spatiotemporal Air Quality Index (AQI) mapping using sparse sensor data, satellite imagery, and various spatiotemporal factors. By leveraging Graph Neural Networks (GNNs), we estimate AQI values at unmonitored locations based on both spatial and temporal dependencies. The framework incorporates a wide range of environmental features, including meteorological data, road networks, points of interest (PoIs), population density, and urban green spaces, which enhance prediction accuracy. We illustrate the use of our approach through a case study in Lahore, Pakistan, where multi-resolution data is used to generate the air quality index map at a fine spatiotemporal scale.

Large language models (LLMs) offer promise in generating educational content, providing instructor feedback, and reducing teacher workload on assessments. While prior studies have focused on studying LLM-powered learning analytics, limited research has examined how effective LLMs are in a bilingual context. In this paper, we study the effectiveness of multilingual large language models (MLLMs) across monolingual (English-only, Spanish-only) and bilingual (Spanglish) student writing. We present a learning analytics use case that details LLM performance in assessing acceptable and unacceptable explanations of Science and Social Science concepts. Our findings reveal a significant bias in the grading performance of pre-trained models for bilingual writing compared to English-only and Spanish-only writing. Following this, we fine-tune open-source MLLMs including Llama 3.1 and Mistral NeMo using synthetic datasets generated in English, Spanish, and Spanglish. Our experiments indicate that the models perform significantly better for all three languages after fine-tuning with bilingual data. This study highlights the potential of enhancing MLLM effectiveness to support authentic language practices amongst bilingual learners. It also aims to illustrate the value of incorporating non-English languages into the design and implementation of language models in education.

Personalized content recommendations have been pivotal to the content experience in digital media from video streaming to social networks. However, several domain specific challenges have held back adoption of recommender systems in news publishing. To address these challenges, we introduce the Ekstra Bladet News Recommendation Dataset (EB-NeRD). The dataset encompasses data from over a million unique users and more than 37 million impression logs from Ekstra Bladet. It also includes a collection of over 125,000 Danish news articles, complete with titles, abstracts, bodies, and metadata, such as categories. EB-NeRD served as the benchmark dataset for the RecSys '24 Challenge, where it was demonstrated how the dataset can be used to address both technical and normative challenges in designing effective and responsible recommender systems for news publishing. The dataset is available at: https://recsys.eb.dk., Comment: 11 pages, 8 tables, 2 figures, RecSys '24

The RecSys Challenge 2024 aims to advance news recommendation by addressing both the technical and normative challenges inherent in designing effective and responsible recommender systems for news publishing. This paper describes the challenge, including its objectives, problem setting, and the dataset provided by the Danish news publishers Ekstra Bladet and JP/Politikens Media Group ("Ekstra Bladet"). The challenge explores the unique aspects of news recommendation, such as modeling user preferences based on behavior, accounting for the influence of the news agenda on user interests, and managing the rapid decay of news items. Additionally, the challenge embraces normative complexities, investigating the effects of recommender systems on news flow and their alignment with editorial values. We summarize the challenge setup, dataset characteristics, and evaluation metrics. Finally, we announce the winners and highlight their contributions. The dataset is available at: https://recsys.eb.dk., Comment: 5 pages, 3 tables, RecSys' 24

Our objective is to investigate the distribution of dust and associated large-scale structures of the Galaxy using optical linear polarization measurements of various open clusters located at different distances in the Galactic anticenter direction. We present R-band linear polarization observations of stars towards five open clusters: Kronberger~1, Berkeley~69, Berkeley~71, Berkeley~19, and King~8 in the anticenter direction. The polarization observations were carried out using AIMPOL instrument mounted on the 104 cm Sampurnanand telescope of ARIES, Nainital, making it the first study to target the polarization observations towards distant clusters ($\sim$6~kpc). We combined the observed polarization data with the distance information from the Gaia space telescope to infer the dust distribution along the line of sight. The variation in the degree of polarization and extinction with distance reveals the presence of multiple dust layers along each cluster direction. In addition, common foreground dust layers detected towards different cluster directions highlight the presence of global features such as spiral arms. Our results show that the dust clouds at 2~kpc towards Berkeley~69 and Berkeley~71 coincide with the Perseus arm, while the dust layer at $\sim$4~kpc towards distant clusters, Berkeley~19 and King~8, indicates the presence of the Outer arm. The large-scale dust distribution obtained by combining our polarization results with the previous polarization studies of nearby open clusters suggests that the anticenter direction is characterized by low extinction, homogeneous dust distribution with somewhat uniform orientation of the plane-of-sky component of the magnetic field along the line of sight. Our study demonstrates the utility of polarization as a tool to study the large-scale dust distribution., Comment: 22 Pages, 14 Figures, 7 Tables, Accepted for publication in Astronomy & Astrophysics Journal

Wide-Area Linear Optical Polarimeter (WALOP)-South is the first wide-field and survey-capacity polarimeter in the optical wavelengths. On schedule for commissioning in 2024, it will be mounted on the 1 m SAAO telescope in Sutherland Observatory, South Africa to undertake the PASIPHAE sky survey. PASIPHAE program will create the first polarimetric sky map in the optical wavelengths, spanning more than 2000 square degrees of the southern Galactic region. The innovative design of WALOP-South will enable it to measure the linear polarization (Stokes parameters $q$ and $u$), in a single exposure, of all sources in a field of view (FoV) of $35\times35$ arcminutes-squared in the SDSS-r broadband and narrowband filters between 500-750 nm with 0.1 % polarization accuracy. The unique goals of the instrument place very stringent systems engineering goals, including on the performance of the optical, polarimetric, optomechanical, and electronic subsystems. All the subsystems have been designed carefully to meet the overall instrument performance goals. As of May 2024, all the instrument optical and mechanical subsystems have been assembled and are currently getting tested and integrated. The complete testing and characterization of the instrument in the lab is expected to be completed by August 2024. In this paper, we will present (a) the design and development of the entire instrument and its major subsystems, focusing on the opto-mechanical design which has not been reported before, and (b) assembly and integration of the instrument in the lab and early results from lab characterization of the instrument., Comment: 20 pages; presented at 2024 SPIE Astronomical Telescopes + Instrumentation conference

Supermassive black holes (SMBHs) can be ejected from their galactic centers due to gravitational wave recoil or the slingshot mechanism following a galaxy merger. If an ejected SMBH retains its inner accretion disk, it may be visible as an off-nuclear active galactic nucleus (AGN). At present, only a handful of offset AGNs that are recoil or slingshot candidates have been found, and none have been robustly confirmed. Compiling a large sample of runaway SMBHs would enable us to constrain the mass and spin evolution of binary SMBHs and study feedback effects of displaced AGNs. We adapt the method of varstrometry -- which was developed for Gaia observations to identify off-center, dual, and lensed AGNs -- in order to quickly identify off-nuclear AGNs in optical survey data by looking for an excess of blue versus red astrometric jitter. We apply this to the Pan-STARRS1 3$\pi$ Survey and report on five new runaway AGN candidates. We focus on ZTF18aajyzfv: a luminous quasar offset by 6.7 $\pm$ 0.2 kpc from an adjacent galaxy at $z$=0.224, and conclude after Keck LRIS spectroscopy and comparison to ASTRID simulation analogs that it is likely a dual AGN. This selection method can be easily adapted to work with data from the soon-to-be commissioned Vera C. Rubin Telescope Legacy Survey of Space and Time (LSST). LSST will have a higher cadence and deeper magnitude limit than Pan-STARRS1, and should permit detection of many more runaway SMBH candidates., Comment: Accepted for publication by ApJ. 13 pages, 12 figures, 3 tables

While there has been remarkable progress recently in the fields of manipulation and locomotion, mobile manipulation remains a long-standing challenge. Compared to locomotion or static manipulation, a mobile system must make a diverse range of long-horizon tasks feasible in unstructured and dynamic environments. While the applications are broad and interesting, there are a plethora of challenges in developing these systems such as coordination between the base and arm, reliance on onboard perception for perceiving and interacting with the environment, and most importantly, simultaneously integrating all these parts together. Prior works approach the problem using disentangled modular skills for mobility and manipulation that are trivially tied together. This causes several limitations such as compounding errors, delays in decision-making, and no whole-body coordination. In this work, we present a reactive mobile manipulation framework that uses an active visual system to consciously perceive and react to its environment. Similar to how humans leverage whole-body and hand-eye coordination, we develop a mobile manipulator that exploits its ability to move and see, more specifically -- to move in order to see and to see in order to move. This allows it to not only move around and interact with its environment but also, choose "when" to perceive "what" using an active visual system. We observe that such an agent learns to navigate around complex cluttered scenarios while displaying agile whole-body coordination using only ego-vision without needing to create environment maps. Results visualizations and videos at https://spin-robot.github.io/, Comment: In CVPR 2024. Website at https://spin-robot.github.io/

Visual neural decoding from EEG has improved significantly due to diffusion models that can reconstruct high-quality images from decoded latents. While recent works have focused on relatively complex architectures to achieve good reconstruction performance from EEG, less attention has been paid to the source of this information. In this work, we attempt to discover EEG features that represent perceptual and semantic visual categories, using a simple pipeline. Notably, the high temporal resolution of EEG allows us to go beyond static semantic maps as obtained from fMRI. We show (a) Training a simple linear decoder from EEG to CLIP latent space, followed by a frozen pre-trained diffusion model, is sufficient to decode images with state-of-the-art reconstruction performance. (b) Mapping the decoded latents back to EEG using a linear encoder isolates CLIP-relevant EEG spatiotemporal features. (c) By using other latent spaces representing lower-level image features, we obtain similar time-courses of texture/hue-related information. We thus use our framework, Perceptogram, to probe EEG signals at various levels of the visual information hierarchy.

Many species of microbes cooperate by producing public goods from which they collectively benefit. However, these populations are under the risk of being taken over by cheating mutants that do not contribute to the pool of public goods. Here we present theoretical findings that address how the social evolution of microbes can be manipulated by external perturbations, to inhibit or promote the fixation of cheaters. To control social evolution, we determine the effects of fluid-dynamical properties such as flow rate or boundary geometry. We also study the social evolutionary consequences of introducing beneficial or harmful chemicals at steady state and in a time dependent fashion. We show that by modulating the flow rate and by applying pulsed chemical signals, we can modulate the spatial structure and dynamics of the population, in a way that can select for more or less cooperative microbial populations., Comment: 16 pages, 5 figures

While there have been significant strides in dexterous manipulation, most of it is limited to benchmark tasks like in-hand reorientation which are of limited utility in the real world. The main benefit of dexterous hands over two-fingered ones is their ability to pickup tools and other objects (including thin ones) and grasp them firmly to apply force. However, this task requires both a complex understanding of functional affordances as well as precise low-level control. While prior work obtains affordances from human data this approach doesn't scale to low-level control. Similarly, simulation training cannot give the robot an understanding of real-world semantics. In this paper, we aim to combine the best of both worlds to accomplish functional grasping for in-the-wild objects. We use a modular approach. First, affordances are obtained by matching corresponding regions of different objects and then a low-level policy trained in sim is run to grasp it. We propose a novel application of eigengrasps to reduce the search space of RL using a small amount of human data and find that it leads to more stable and physically realistic motion. We find that eigengrasp action space beats baselines in simulation and outperforms hardcoded grasping in real and matches or outperforms a trained human teleoperator. Results visualizations and videos at https://dexfunc.github.io/, Comment: In CoRL 2023. Website at https://dexfunc.github.io/

Dust is a ubiquitous component in our Galaxy. It accounts for only $1\%$ mass of the ISM but still is an essential part of the Galaxy. It affects our view of the Galaxy by obscuring the starlight at shorter wavelengths and re-emitting in longer wavelengths. Studying the dust distribution in the Galaxy at longer wavelengths may cause discrepancies due to distance ambiguity caused by unknown Galactic potential. However, another aspect of dust, i.e., the polarisation of the background starlight, when combined with distance information, will help to give direct observational evidence of the number of dust clouds encountered in the line of sight. We observed 15 open clusters distributed at increasing distances in three lines of sight using two Indian national facilities. The measured polarisation results used to scrutinize the dust distribution and orientation of the local plane of sky magnetic fields towards selected directions. The analysis of the stars observed towards the distant cluster King 8 cluster shows two foreground layers at a distance of $\sim 500$ pc and $\sim$ 3500 pc. Similar analysis towards different clusters also results in multiple dust layers., Comment: 11 pages, 5 figures, 1 table Accepted for publication in the Bulletin of Li\`ege Royal Society of Sciences as a part of 3rd Belgo-Indian Network for Astronomy and Astrophysics (BINA) workshop, 22-24 March 2023

Our study aims to investigate the outer disc structure of the Milky Way galaxy using the red clump (RC) stars. We analysed the distribution of the largest sample of RC stars to date, homogeneously covering the entire Galactic plane in the range of $40^\circ \le \ell \le 340^\circ$ and $-10^\circ \le b \le +10^\circ$. This sample allows us to model the RC star distribution in the Galactic disc to better constrain the properties of the flare and warp of the Galaxy. Our results show that the scale length of the old stellar disc weakly depends on azimuth, with an average value of $1.95 \pm0.26$ kpc. On the other hand, a significant disc flaring is detected, where the scale height of the disc increases from 0.38 kpc in the solar neighbourhood to $\sim 2.2$ kpc at R $\approx 15$ kpc. The flare exhibits a slight asymmetry, with $\sim 1$ kpc more scale height below the Galactic plane as compared to the Northern flare. We also confirm the warping of the outer disc, which can be modelled with $Z_w = (0.0057 \pm 0.0050)~ [R-(7358 \pm 368) (pc)]^{1.40 \pm 0.09} \sin(\phi - (-2^\circ.03 \pm 0^\circ .18))$. Our analysis reveals a noticeable north-south asymmetry in the warp, with a greater amplitude observed in the southern direction compared to the northern. Comparing our findings with younger tracers from the literature, we observe an age dependency of both the flare and warp. An increase in flare strength with age suggests the secular evolution of the disc as the preferred mechanism for forming the flare. The increase of the maximum warp amplitude with age indicates that the warp dynamics could be the possible cause of the variation in the warp properties with age., Comment: 11 pages, 12 figures, 4 tables Accepted for publication in MNRAS

We study the integral Brauer--Manin obstruction for affine diagonal cubic surfaces, which we employ to construct the first counterexamples to the integral Hasse principle in this setting. We then count in three natural ways how such counterexamples are distributed across the family of affine diagonal cubic surfaces and how often such surfaces satisfy integral strong approximation off $\infty$., Comment: 38 pages

If $X$ is a singular del Pezzo surface of degree $d$ over a finite field $\mathbb{F}_{q}$ with only rational double point singularities, does there always exist a smooth $\mathbb{F}_{q}$-point on $X$? We show that this is true for $d\geq 3$ and give counterexamples in the case of $d=2$.

This paper considers the problem of extracting building footprints from satellite imagery -- a task that is critical for many urban planning and decision-making applications. While recent advancements in deep learning have made great strides in automated detection of building footprints, state-of-the-art methods available in existing literature often generate erroneous results for areas with densely connected buildings. Moreover, these methods do not incorporate the context of neighborhood images during training thus generally resulting in poor performance at image boundaries. In light of these gaps, we propose a novel Tuning Fork Network (TFNet) design for deep semantic segmentation that not only performs well for widely-spaced building but also has good performance for buildings that are closely packed together. The novelty of TFNet architecture lies in a a single encoder followed by two parallel decoders to separately reconstruct the building footprint and the building edge. In addition, the TFNet design is coupled with a novel methodology of incorporating neighborhood information at the tile boundaries during the training process. This methodology further improves performance, especially at the tile boundaries. For performance comparisons, we utilize the SpaceNet2 and WHU datasets, as well as a dataset from an area in Lahore, Pakistan that captures closely connected buildings. For all three datasets, the proposed methodology is found to significantly outperform benchmark methods.

In variational inference, the benefits of Bayesian models rely on accurately capturing the true posterior distribution. We propose using neural samplers that specify implicit distributions, which are well-suited for approximating complex multimodal and correlated posteriors in high-dimensional spaces. Our approach introduces novel bounds for approximate inference using implicit distributions by locally linearising the neural sampler. This is distinct from existing methods that rely on additional discriminator networks and unstable adversarial objectives. Furthermore, we present a new sampler architecture that, for the first time, enables implicit distributions over tens of millions of latent variables, addressing computational concerns by using differentiable numerical approximations. We empirically show that our method is capable of recovering correlations across layers in large Bayesian neural networks, a property that is crucial for a network's performance but notoriously challenging to achieve. To the best of our knowledge, no other method has been shown to accomplish this task for such large models. Through experiments in downstream tasks, we demonstrate that our expressive posteriors outperform state-of-the-art uncertainty quantification methods, validating the effectiveness of our training algorithm and the quality of the learned implicit approximation., Comment: 10 pages and appendix, 9 figures, 7 tables

Logistic hubs play a pivotal role in the last-mile delivery distance; even a slight increment in distance negatively impacts the business of the e-commerce industry while also increasing its carbon footprint. The growth of this industry, particularly after Covid-19, has further intensified the need for optimized allocation of resources in an urban environment. In this study, we use a hybrid approach to optimize the placement of logistic hubs. The approach sequentially employs different techniques. Initially, delivery points are clustered using K-Means in relation to their spatial locations. The clustering method utilizes road network distances as opposed to Euclidean distances. Non-road network-based approaches have been avoided since they lead to erroneous and misleading results. Finally, hubs are located using the P-Median method. The P-Median method also incorporates the number of deliveries and population as weights. Real-world delivery data from Muller and Phipps (M&P) is used to demonstrate the effectiveness of the approach. Serving deliveries from the optimal hub locations results in the saving of 815 (10%) meters per delivery.

Any policy-level decision-making procedure and academic research involving the optimum use of resources for development and planning initiatives depends on accurate population density statistics. The current cutting-edge datasets offered by WorldPop and Meta do not succeed in achieving this aim for developing nations like Pakistan; the inputs to their algorithms provide flawed estimates that fail to capture the spatial and land-use dynamics. In order to precisely estimate population counts at a resolution of 30 meters by 30 meters, we use an accurate built settlement mask obtained using deep segmentation networks and satellite imagery. The Points of Interest (POI) data is also used to exclude non-residential areas.

Profitant du travail de pr\'ec\'edent d'Harpaz nous utilisons la m\'ethode de descente-fibration de Swinnerton-Dyer pour \'etudier les points int\'egraux sur des surfaces affines qui sont des fibration de tores de norme 1 sur $\mathbb{A}^{1}$. Taking advantage of previous work of Harpaz we use Swinnerton-Dyer's descent-fibration method to study integral points on affine surfaces which are fibrations of norm 1 tori over $\mathbb{A}^{1}$.

In this article, we consider the problem of outlier-robust state estimation where the measurement noise can be correlated. Outliers in data arise due to many reasons like sensor malfunctioning, environmental behaviors, communication glitches, etc. Moreover, noise correlation emerges in several real-world applications e.g. sensor networks, radar data, GPS-based systems, etc. We consider these effects in system modeling which is subsequently used for inference. We employ the Expectation-Maximization (EM) framework to derive both outlier-resilient filtering and smoothing methods, suitable for online and offline estimation respectively. The standard Gaussian filtering and the Gaussian Rauch-Tung-Striebel (RTS) smoothing results are leveraged to devise the estimators. In addition, Bayesian Cramer-Rao Bounds (BCRBs) for a filter and a smoother which can perfectly detect and reject outliers are presented. These serve as useful theoretical benchmarks to gauge the error performance of different estimators. Lastly, different numerical experiments, for an illustrative target tracking application, are carried out that indicate performance gains compared to similarly engineered state-of-the-art outlier-rejecting state estimators. The advantages are in terms of simpler implementation, enhanced estimation quality, and competitive computational performance., Comment: 13 pages, 9 figures

This study uses 2019 data from the University and College Academic Staff System to examine gender differences in tenure status among faculty in Canadian universities. It also uses the Survey of Postsecondary Faculty and Researchers to examine feelings of fairness in hiring and promotions. (1) In 2019, almost 6 in 10 full-time university faculty members in Canada were men. Male university faculty are older and more likely to be full professors than women; (2) Female faculty are less likely to be in tenured positions than male faculty. In 2019, less than two thirds (63%) of female faculty held such positions, compared with three quarters (75%) of their male counterparts; (3) Once multiple factors associated with tenure were considered (including time since completion of the highest degree and the highest level of education), the gender gap in tenure was reduced. However, women remained less likely to be in tenured positions than men (68% versus 71%); (4) Female faculty were less likely to feel that hiring was fair and equitable at their institution. Specifically, around 20% of women surveyed disagreed or strongly disagreed that hiring was fair and equitable at their institution, compared with 12% of men; (5) Women (23%) were also more likely than men (14%) to state that they disagreed or strongly disagreed that promotions were fair and equitable at their institution; and (6) Other groups that were less likely to feel that hiring was fair and equitable included faculty belonging to racialized groups, faculty members with a disability and sexual minority female faculty. In addition, faculty members with a disability and women who belong to racialized groups were less likely to feel promotions were fair and equitable. [This study was funded by Women and Gender Equality Canada (WAGE).]