Search

Image classification models, including convolutional neural networks (CNNs), perform well on a variety of classification tasks but struggle under conditions of partial occlusion, i.e., conditions in which objects are partially covered from the view of a camera. Methods to improve performance under occlusion, including data augmentation, part-based clustering, and more inherently robust architectures, including Vision Transformer (ViT) models, have, to some extent, been evaluated on their ability to classify objects under partial occlusion. However, evaluations of these methods have largely relied on images containing artificial occlusion, which are typically computer-generated and therefore inexpensive to label. Additionally, methods are rarely compared against each other, and many methods are compared against early, now outdated, deep learning models. We contribute the Image Recognition Under Occlusion (IRUO) dataset, based on the recently developed Occluded Video Instance Segmentation (OVIS) dataset (arXiv:2102.01558). IRUO utilizes real-world and artificially occluded images to test and benchmark leading methods' robustness to partial occlusion in visual recognition tasks. In addition, we contribute the design and results of a human study using images from IRUO that evaluates human classification performance at multiple levels and types of occlusion. We find that modern CNN-based models show improved recognition accuracy on occluded images compared to earlier CNN-based models, and ViT-based models are more accurate than CNN-based models on occluded images, performing only modestly worse than human accuracy. We also find that certain types of occlusion, including diffuse occlusion, where relevant objects are seen through "holes" in occluders such as fences and leaves, can greatly reduce the accuracy of deep recognition models as compared to humans, especially those with CNN backbones.

Artificial Intelligence (AI) technologies have profoundly transformed the field of remote sensing, revolutionizing data collection, processing, and analysis. Traditionally reliant on manual interpretation and task-specific models, remote sensing has been significantly enhanced by the advent of foundation models--large-scale, pre-trained AI models capable of performing a wide array of tasks with unprecedented accuracy and efficiency. This paper provides a comprehensive survey of foundation models in the remote sensing domain, covering models released between June 2021 and June 2024. We categorize these models based on their applications in computer vision and domain-specific tasks, offering insights into their architectures, pre-training datasets, and methodologies. Through detailed performance comparisons, we highlight emerging trends and the significant advancements achieved by these foundation models. Additionally, we discuss the technical challenges, practical implications, and future research directions, addressing the need for high-quality data, computational resources, and improved model generalization. Our research also finds that pre-training methods, particularly self-supervised learning techniques like contrastive learning and masked autoencoders, significantly enhance the performance and robustness of foundation models in remote sensing tasks such as scene classification, object detection, and other applications. This survey aims to serve as a resource for researchers and practitioners by providing a panorama of advances and promising pathways for continued development and application of foundation models in remote sensing.

The auxiliary field diffusion Monte Carlo method uses imaginary-time projection techniques to accurately solve the ground-state wave function of atomic nuclei and infinite nuclear matter. In this work, we present a novel representation of the imaginary-time propagator based on restricted Boltzmann machines. We test its accuracy against the routinely employed Hubbard-Stratonovich transformations by evaluating ground-state energies and single-particle densities of selected light nuclei. This analysis paves the way for incorporating more realistic nuclear potentials in the auxiliary field diffusion Monte Carlo method, including isospin-dependent spin-orbit terms and cubic spin-isospin operators, which characterize accurate phenomenological and chiral effective field theory Hamiltonians., Comment: 8 pages, 3 figures

The time variability and spectra of directly imaged companions provide insight into their physical properties and atmospheric dynamics. We present follow-up R~40 spectrophotometric monitoring of red companion HD 1160 B at 2.8-4.2 $\mu$m using the double-grating 360{\deg} vector Apodizing Phase Plate (dgvAPP360) coronagraph and ALES integral field spectrograph on the Large Binocular Telescope Interferometer. We use the recently developed technique of gvAPP-enabled differential spectrophotometry to produce differential light curves for HD 1160 B. We reproduce the previously reported ~3.2 h periodic variability in archival data, but detect no periodic variability in new observations taken the following night with a similar 3.5% level precision, suggesting rapid evolution in the variability of HD 1160 B. We also extract complementary spectra of HD 1160 B for each night. The two are mostly consistent, but the companion appears fainter on the second night between 3.0-3.2 $\mu$m. Fitting models to these spectra produces different values for physical properties depending on the night considered. We find an effective temperature T$_{\text{eff}}$ = 2794$^{+115}_{-133}$ K on the first night, consistent with the literature, but a cooler T$_{\text{eff}}$ = 2279$^{+79}_{-157}$ K on the next. We estimate the mass of HD 1160 B to be 16-81 M$_{\text{Jup}}$, depending on its age. We also present R = 50,000 high-resolution optical spectroscopy of host star HD 1160 A obtained simultaneously with the PEPSI spectrograph. We reclassify its spectral type to A1 IV-V and measure its projected rotational velocity v sin i = 96$^{+6}_{-4}$ km s$^{-1}$. We thus highlight that gvAPP-enabled differential spectrophotometry can achieve repeatable few percent level precision and does not yet reach a systematic noise floor, suggesting greater precision is achievable with additional data or advanced detrending techniques., Comment: 22 pages, 12 figures, published in MNRAS; typos corrected, references updated

We show that direct laser writing in aqueous silver nitrate with a $\lambda$ = 1030 nm femtosecond laser results in deposition of a mixture of silver oxide and silver, in contrast to the pure silver deposition previously reported with 780 nm femtosecond direct laser writing. However, adding photoinitiator prevents silver oxide formation in a concentration-dependent manner. As a result, the resistivity of the material can also be controlled by photoinitiator concentration with resistivity being reduced from approximately 9e-3 $\Omega$m to 3e-7 $\Omega$m. Silver oxide peaks dominate the X-ray diffraction spectra when no photoinitiator is present, while the peaks disappear with photoinitiator concentrations above 0.05 wt%. A THz polarizer and metamaterial are printed as a demonstration of silver oxide printing.

Large language models (LLMs) such as ChatGPT, Gemini, LlaMa, and Claude are trained on massive quantities of text parsed from the internet and have shown a remarkable ability to respond to complex prompts in a manner often indistinguishable from humans. We present a LLM fine-tuned on up to 40,000 data that can predict electromagnetic spectra over a range of frequencies given a text prompt that only specifies the metasurface geometry. Results are compared to conventional machine learning approaches including feed-forward neural networks, random forest, linear regression, and K-nearest neighbor (KNN). Remarkably, the fine-tuned LLM (FT-LLM) achieves a lower error across all dataset sizes explored compared to all machine learning approaches including a deep neural network. We also demonstrate the LLM's ability to solve inverse problems by providing the geometry necessary to achieve a desired spectrum. LLMs possess some advantages over humans that may give them benefits for research, including the ability to process enormous amounts of data, find hidden patterns in data, and operate in higher-dimensional spaces. We propose that fine-tuning LLMs on large datasets specific to a field allows them to grasp the nuances of that domain, making them valuable tools for research and analysis.

The ability for quantum and conventional networks to operate in the same optical fibers would aid the deployment of quantum network technology on a large scale. Quantum teleportation is a fundamental operation in quantum networking, but has yet to be demonstrated in fibers populated with high-power conventional optical signals. Here we report to the best of our knowledge the first demonstration of quantum teleportation over fibers carrying conventional telecommunications traffic. Quantum state transfer is achieved over a 30.2-km fiber carrying 400-Gbps C-band classical traffic with a Bell state measurement performed at the fiber midpoint. To protect quantum fidelity from spontaneous Raman scattering noise, we use optimal O-band quantum channels, narrow spectro-temporal filtering, and multi-photon coincidence detection. Fidelity is shown to be well maintained with an elevated C-band classical power of 18.7 dBm, which could support multiple classical channels totaling many terabits/s aggregate data rates. These results show the feasibility of advanced quantum and classical network applications operating within a unified fiber infrastructure.

We analyze the static response to kinetic perturbations of nonequilibrium steady states that can be modeled as diffusions. We demonstrate that kinetic response is purely a nonequilibirum effect, measuring the degree to which the Fluctuation-Dissipation Theorem is violated out of equilibrium. For driven diffusions in a flat landscape, we further demonstrate that such response is constrained by the strength of the nonequilibrium driving via quantitative inequalities., Comment: 7 pages, 3 figures

Preceding a core-collapse supernova, various processes produce an increasing amount of neutrinos of all flavors characterized by mounting energies from the interior of massive stars. Among them, the electron antineutrinos are potentially detectable by terrestrial neutrino experiments such as KamLAND and Super-Kamiokande via inverse beta decay interactions. Once these pre-supernova neutrinos are observed, an early warning of the upcoming core-collapse supernova can be provided. In light of this, KamLAND and Super-Kamiokande, both located in the Kamioka mine in Japan, have been monitoring pre-supernova neutrinos since 2015 and 2021, respectively. Recently, we performed a joint study between KamLAND and Super-Kamiokande on pre-supernova neutrino detection. A pre-supernova alert system combining the KamLAND detector and the Super-Kamiokande detector was developed and put into operation, which can provide a supernova alert to the astrophysics community. Fully leveraging the complementary properties of these two detectors, the combined alert is expected to resolve a pre-supernova neutrino signal from a 15 M$_{\odot}$ star within 510 pc of the Earth, at a significance level corresponding to a false alarm rate of no more than 1 per century. For a Betelgeuse-like model with optimistic parameters, it can provide early warnings up to 12 hours in advance., Comment: Resubmitted to ApJ. 22 pages, 16 figures, for more information about the combined pre-supernova alert system, see https://www.lowbg.org/presnalarm/

Mechanical force is an essential feature for many physical and biological processes.1-12 Remote measurement of mechanical signals with high sensitivity and spatial resolution is needed for diverse applications, including robotics,13 biophysics,14-20 energy storage,21-24 and medicine.25-27 Nanoscale luminescent force sensors excel at measuring piconewton forces,28-32 while larger sensors have proven powerful in probing micronewton forces.33,34 However, large gaps remain in the force magnitudes that can be probed remotely from subsurface or interfacial sites, and no individual, non-invasive sensor is capable of measuring over the large dynamic range needed to understand many systems.35,36 Here, we demonstrate Tm3+-doped avalanching nanoparticle37 force sensors that can be addressed remotely by deeply penetrating near-infrared (NIR) light and can detect piconewton to micronewton forces with a dynamic range spanning more than four orders of magnitude. Using atomic force microscopy coupled with single-nanoparticle optical spectroscopy, we characterize the mechanical sensitivity of the photon avalanching process and reveal its exceptional force responsiveness. By manipulating the Tm3+ concentrations and energy transfer within the nanosensors, we demonstrate different optical force-sensing modalities, including mechanobrightening and mechanochromism. The adaptability of these nanoscale optical force sensors, along with their multiscale sensing capability, enable operation in the dynamic and versatile environments present in real-world, complex structures spanning biological organisms to nanoelectromechanical systems (NEMS).

Image-guided navigation of catheter devices to anatomic targets within large 3D cardiac chambers and vessels is challenging in the interventional setting due to the limitations of a conventional 2D x-ray projection format. Scanning-beam digital x-ray (SBDX) is a low-dose inverse geometry x-ray fluoroscopy technology capable of real-time 3D catheter tracking. SBDX performs rapid tomosynthesis using an electronically scanned multisource x-ray tube and photon-counting detector mounted to a C-arm gantry. However, SBDX currently lacks the ability to perform volumetric computed tomography from a rotational C-arm scan. This work develops a novel volumetric CT capability for the SBDX platform, termed C-arm inverse geometry CT, suitable for rotational scans of the beating heart. The work is divided into three tasks: development of image reconstruction algorithms, implementation on the SBDX hardware, and performance assessment for the example task of 3D cardiac chamber mapping. SBDX-CT data acquisition is performed by simultaneous x-ray source scanning at 15 scan/s and C-arm rotation over a 190 degree short-scan arc in 13.4 seconds. An iterative reconstruction method was developed to accommodate fully truncated projections and data inconsistency resulting from cardiac motion during rotation. Hardware implementation included development of a C-arm angle measurement method, development of a geometric calibration method to account for non-ideal C-arm rotations, and detector response nonlinearity correction. The geometric calibration procedure mitigated artifacts from C-arm deflection during rotation. SBDX-CT image quality was evaluated in phantom studies. Feasibility of in-vivo SBDX-CT was demonstrated in a porcine model.

We introduce a novel method for studying systematic trends in nuclear reaction data using generative adversarial networks. Libraries of nuclear cross section evaluations exhibit intricate systematic trends across the nuclear landscape, and predictive models capable of reproducing and analyzing these trends are valuable for many applications. We have developed a predictive model using deep generative adversarial networks to learn trends from the inelastic neutron scattering channel of TENDL for even-even nuclei. The system predicts cross sections based on adding/subtracting particles to/from the target nucleus. It can thus help identify cross sections that break from expected trends and predict beyond the limit of current experiments. Our model can produce good predictions for cross section curves for many nuclides, and it is most robust near the line of stability. We also create an ensemble of predictions to leverage different correlations and estimate model uncertainty. This research marks an important first step in computer generation of nuclear cross-section libraries., Comment: New version uploaded after rewriting several sections for clarity and one new figure added

The unprecedented medium-resolution (R~1500-3500) near- and mid-infrared (1-18um) spectrum provided by JWST for the young (140+/-20Myr) low-mass (12-20MJup) L-T transition (L7) companion VHS1256b gives access to a catalogue of molecular absorptions. In this study, we present a comprehensive analysis of this dataset utilizing a forward modelling approach, applying our Bayesian framework, ForMoSA. We explore five distinct atmospheric models to assess their performance in estimating key atmospheric parameters: Teff, log(g), [M/H], C/O, gamma, fsed, and R. Our findings reveal that each parameter's estimate is significantly influenced by factors such as the wavelength range considered and the model chosen for the fit. This is attributed to systematic errors in the models and their challenges in accurately replicating the complex atmospheric structure of VHS1256b, notably the complexity of its clouds and dust distribution. To propagate the impact of these systematic uncertainties on our atmospheric property estimates, we introduce innovative fitting methodologies based on independent fits performed on different spectral windows. We finally derived a Teff consistent with the spectral type of the target, considering its young age, which is confirmed by our estimate of log(g). Despite the exceptional data quality, attaining robust estimates for chemical abundances [M/H] and C/O, often employed as indicators of formation history, remains challenging. Nevertheless, the pioneering case of JWST's data for VHS1256b has paved the way for future acquisitions of substellar spectra that will be systematically analyzed to directly compare the properties of these objects and correct the systematics in the models., Comment: 32 pages, 16 figures, 6 tables, 2 appendices

We present aperture masking interferometry (AMI) observations of the star HIP 65426 at $3.8\,\rm{\mu m}$ as a part of the \textit{JWST} Direct Imaging Early Release Science (ERS) program obtained using the Near Infrared Imager and Slitless Spectrograph (NIRISS) instrument. This mode provides access to very small inner working angles (even separations slightly below the Michelson limit of ${}0.5\lambda/D$ for an interferometer), which are inaccessible with the classical inner working angles of the \textit{JWST} coronagraphs. When combined with \textit{JWST}'s unprecedented infrared sensitivity, this mode has the potential to probe a new portion of parameter space across a wide array of astronomical observations. Using this mode, we are able to achieve a contrast of $\Delta m_{F380M}{\sim }7.8$\,mag relative to the host star at a separation of ${\sim}0.07\arcsec$ but detect no additional companions interior to the known companion HIP\,65426\,b. Our observations thus rule out companions more massive than $10{-}12\,\rm{M\textsubscript{Jup}}$ at separations ${\sim}10{-}20\,\rm{au}$ from HIP\,65426, a region out of reach of ground or space-based coronagraphic imaging. These observations confirm that the AMI mode on \textit{JWST} is sensitive to planetary mass companions orbiting at the water frost line, even for more distant stars at $\sim$100\,pc. This result will allow the planning and successful execution of future observations to probe the inner regions of nearby stellar systems, opening essentially unexplored parameter space., Comment: 15 pages, 9 figures, submitted to ApJ Letters

We present a performance analysis for the aperture masking interferometry (AMI) mode on board the James Webb Space Telescope Near Infrared Imager and Slitless Spectrograph (JWST/NIRISS). Thanks to self-calibrating observables, AMI accesses inner working angles down to and even within the classical diffraction limit. The scientific potential of this mode has recently been demonstrated by the Early Release Science (ERS) 1386 program with a deep search for close-in companions in the HIP 65426 exoplanetary system. As part of ERS 1386, we use the same data set to explore the random, static, and calibration errors of NIRISS AMI observables. We compare the observed noise properties and achievable contrast to theoretical predictions. We explore possible sources of calibration errors and show that differences in charge migration between the observations of HIP 65426 and point-spread function calibration stars can account for the achieved contrast curves. Lastly, we use self-calibration tests to demonstrate that with adequate calibration NIRISS F380M AMI can reach contrast levels of $\sim9-10$ mag at $\gtrsim \lambda/D$. These tests lead us to observation planning recommendations and strongly motivate future studies aimed at producing sophisticated calibration strategies taking these systematic effects into account. This will unlock the unprecedented capabilities of JWST/NIRISS AMI, with sensitivity to significantly colder, lower-mass exoplanets than lower-contrast ground-based AMI setups, at orbital separations inaccessible to JWST coronagraphy., Comment: 20 pages, 12 figures, accepted to Astrophysical Journal Letters

We summarize significant classical results on (in)determinacy of measures in terms of their finite positive integer order moments. Well-known is the role of the smallest eigenvalues of Hankel matrices, starting from Hamburger's results a century ago and ending with the great progress made only in recent times by C. Berg and collaborators. We describe here known results containing necessary and sufficient conditions for moment (in)determinacy in both Hamburger and Stieltjes moment problems. In our exposition we follow an approach different from that commonly used. There are novelties well complementing the existing theory. Among them are: (a) to emphasize on the geometric interpretation of the indeterminacy conditions; (b) exploit fine properties of the eigenvalues of perturbed symmetric matrices allowing to derive new lower bounds for the smallest eigenvalues of Hankel matrices; these bounds are used for concluding indeterminacy; (c) provide new arguments to confirm classical results; (d) give new numerical illustrations involving commonly used probability distributions., Comment: 23 pages, 5 figures. This version corrects 1 typo and 2 inaccuracies

Translational research requires data at multiple scales of biological organization. Advancements in sequencing and multi-omics technologies have increased the availability of these data, but researchers face significant integration challenges. Knowledge graphs (KGs) are used to model complex phenomena, and methods exist to construct them automatically. However, tackling complex biomedical integration problems requires flexibility in the way knowledge is modeled. Moreover, existing KG construction methods provide robust tooling at the cost of fixed or limited choices among knowledge representation models. PheKnowLator (Phenotype Knowledge Translator) is a semantic ecosystem for automating the FAIR (Findable, Accessible, Interoperable, and Reusable) construction of ontologically grounded KGs with fully customizable knowledge representation. The ecosystem includes KG construction resources (e.g., data preparation APIs), analysis tools (e.g., SPARQL endpoints and abstraction algorithms), and benchmarks (e.g., prebuilt KGs and embeddings). We evaluated the ecosystem by systematically comparing it to existing open-source KG construction methods and by analyzing its computational performance when used to construct 12 large-scale KGs. With flexible knowledge representation, PheKnowLator enables fully customizable KGs without compromising performance or usability.

Limits on a system's response to external perturbations inform our understanding of how physical properties can be shaped by microscopic characteristics. Here, we derive constraints on the steady-state nonequilibrium response of physical observables in terms of the topology of the microscopic state space and the strength of thermodynamic driving. Notably, evaluation of these limits requires no kinetic information beyond the state-space structure. When applied to models of receptor binding, we find that sensitivity is bounded by the steepness of a Hill function with a Hill coefficient enhanced by the chemical driving beyond the structural equilibrium limit., Comment: 25 pages, 13 figures

We propose and investigate a method for identifying timescales of dissipation in nonequilibrium steady states modeled as discrete-state Markov jump processes. The method is based on how the irreversibility-measured by the statistical breaking of time-reversal symmetry-varies under temporal coarse-graining. We observe a sigmoidal-like shape of the irreversibility as a function of the coarse-graining time whose functional form we derive for systems with a fast driven transition. This theoretical prediction allows us to develop a method for estimating the dissipative time scale from time-series data by fitting estimates of the irreversibility to our predicted functional form. We further analyze the accuracy and statistical fluctuations of this estimate., Comment: 15 pages, 4 figures