Search

We reveal a hidden aspect of scale-invariant actions that allow matter to couple with a dilaton field: $\ $the dynamics of the dilaton can be used to induce variations in Planck's quantum of action $\hbar$ and the speed of light $c$. This mechanism for generating variable $\hbar$ and $c$ in $\textit{curved}$ spacetimes offers novel insights into the origin of late-time cosmic acceleration, bypassing the need for dark energy., Comment: 6 pages. No figure

While procedures prevail on the European market for the greater good of its citizens, it might be daunting when trying to introduce a product, whether innovative or not. In the current world, Cyber-Physical Systems (CPSs) are ubiquitous in our daily lives. Cars can provide intrusive assistance as they can brake or turn wheels on their own, buildings are getting smarter to optimize energy consumption, smart cities are emerging to facilitate information sharing and orchestrate the response to emergency situations, etc. As the presence of such tools will grow in the coming years and people will rely even more on CPSs, we certainly need to ensure that they are safe and reliable for users or everybody else, which is why regulations are so important. However, compliance should not act as a barrier to new actors coming to the European market. Nor should it prevent current actors from keeping systems deemed compliant when introduced while obsolete at the time they are used. While the individual elements we point out might not bring novelty in the various research areas we cover (EU policies, requirements engineering, business engineering, and software engineering), this paper identifies the challenges related to building and testing a CPS with respect to applicable laws and discusses the difficulty of automating the response to those challenges, such as finding a relevant legal text, paying for mentioned materials or identifying the level of compliance to a legal text. Our analysis of the holistic context when considering the compliance testing of CPS provides an overview enabling more effective decision-making as well., Comment: 16 pages, 5 figures, IEEEtran

In image-assisted minimally invasive surgeries (MIS), understanding surgical scenes is vital for real-time feedback to surgeons, skill evaluation, and improving outcomes through collaborative human-robot procedures. Within this context, the challenge lies in accurately detecting, segmenting, and estimating the depth of surgical scenes depicted in high-resolution images, while simultaneously reconstructing the scene in 3D and providing segmentation of surgical instruments along with detection labels for each instrument. To address this challenge, a novel Multi-Task Learning (MTL) network is proposed for performing these tasks concurrently. A key aspect of this approach involves overcoming the optimization hurdles associated with handling multiple tasks concurrently by integrating a Adversarial Weight Update into the MTL framework, the proposed MTL model achieves 3D reconstruction through the integration of segmentation, depth estimation, and object detection, thereby enhancing the understanding of surgical scenes, which marks a significant advancement compared to existing studies that lack 3D capabilities. Comprehensive experiments on the EndoVis2018 benchmark dataset underscore the adeptness of the model in efficiently addressing all three tasks, demonstrating the efficacy of the proposed techniques., Comment: 1. Notation Update: Added * for equal contribution, ensuring proper attribution. 2. Subsection Fix: Removed the `subsection` tag for Section 3.1 (no 3.2 existed), maintaining content but fixing hierarchy. 3. Text Additions: Added lines in Section 5 and Subsection 4.2 for clarity, with references for better context

Computational multi-scale pandemic modelling remains a major and timely challenge. Here we identify specific requirements for a new class of pandemic models operating across three scales: (1) rapid pathogen evolution, punctuated by emergence of new variants, (2) human interactions within a heterogeneous population, and (3) public health responses which constrain individual actions to control the disease transmission. We then present a pandemic modelling framework satisfying these requirements and capable of simulating multi-scale dynamic feedback loops. The developed framework comprises a stochastic agent-based model of pandemic spread, coupled with a phylodynamic model of the within-host pathogen evolution. It is validated with a case study, modelling a rapid punctuated evolution of SARS-CoV-2, based on global and contemporary genomic surveillance data, during the COVID-19 transmission within a large heterogeneous population. We demonstrate that the model captures the essential features of the COVID-19 pandemic and the novel coronavirus evolution, while retaining computational tractability and scalability., Comment: 52 pages, 33 figures, 4 tables

Beyond the success of Contrastive Language-Image Pre-training (CLIP), recent trends mark a shift toward exploring the applicability of lightweight vision-language models for resource-constrained scenarios. These models often deliver suboptimal performance when relying solely on a single image-text contrastive learning objective, spotlighting the need for more effective training mechanisms that guarantee robust cross-modal feature alignment. In this work, we propose CLIP-PING: Contrastive Language-Image Pre-training with Proximus Intrinsic Neighbors Guidance, a simple and efficient training paradigm designed to boost the performance of lightweight vision-language models with minimal computational overhead and lower data demands. CLIP-PING bootstraps unimodal features extracted from arbitrary pre-trained encoders to obtain intrinsic guidance of proximus neighbor samples, i.e., nearest-neighbor (NN) and cross nearest-neighbor (XNN). We find that extra contrastive supervision from these neighbors substantially boosts cross-modal alignment, enabling lightweight models to learn more generic features with rich semantic diversity. Extensive experiments reveal that CLIP-PING notably surpasses its peers in zero-shot generalization and cross-modal retrieval tasks. Specifically, a 5.5% gain on zero-shot ImageNet1K with 10.7% (I2T) and 5.7% (T2I) on Flickr30K, compared to the original CLIP when using ViT-XS image encoder trained on 3 million (image, text) pairs. Moreover, CLIP-PING showcases strong transferability under the linear evaluation protocol across several downstream tasks., Comment: 15 pages, 4 figures, 20 tables

Graph-based machine learning methods are useful tools in the identification and prediction of variation in genetic data. In particular, the comprehension of phenotypic effects at the cellular level is an accelerating research area in pharmacogenomics. In this article, a novel graph theoretic approach is proposed to infer a co-occurrence network from 16S microbiome data. The approach is specialised to handle datasets containing a small number of samples. Small datasets exacerbate the significant challenges faced by biological data, which exhibit properties such as sparsity, compositionality, and complexity of interactions. Methodologies are also proposed to enrich and statistically filter the inferred networks. The utility of the proposed method lies in that it extracts an informative network from small sampled data that is not only feature-rich, but also biologically meaningful and statistically significant. Although specialised for small data sets, which are abundant, it can be generally applied to any small-sampled dataset, and can also be extended to integrate multi-omics data. The proposed methodology is tested on a data set of chickens vaccinated against and challenged by the protozoan parasite Eimeria tenella. The raw genetic reads are processed, and networks inferred to describe the ecosystems of the chicken intestines under three different stages of disease progression. Analysis of the expression of network features derive biologically intuitive conclusions from purely statistical methods. For example, there is a clear evolution in the distribution of node features in line with the progression of the disease. The distributions also reveal clusters of species interacting mutualistically and parasitically, as expected. Moreover, a specific sub-network is found to persist through all experimental conditions, representative of a persistent microbiome., Comment: 12 pages, 7 figures, under review for a special issue of ACM/IEEE TCBB journal

In recent years, the rise of machine learning (ML) in cybersecurity has brought new challenges, including the increasing threat of backdoor poisoning attacks on ML malware classifiers. For instance, adversaries could inject malicious samples into public malware repositories, contaminating the training data and potentially misclassifying malware by the ML model. Current countermeasures predominantly focus on detecting poisoned samples by leveraging disagreements within the outputs of a diverse set of ensemble models on training data points. However, these methods are not suitable for scenarios where Machine Learning-as-a-Service (MLaaS) is used or when users aim to remove backdoors from a model after it has been trained. Addressing this scenario, we introduce PBP, a post-training defense for malware classifiers that mitigates various types of backdoor embeddings without assuming any specific backdoor embedding mechanism. Our method exploits the influence of backdoor attacks on the activation distribution of neural networks, independent of the trigger-embedding method. In the presence of a backdoor attack, the activation distribution of each layer is distorted into a mixture of distributions. By regulating the statistics of the batch normalization layers, we can guide a backdoored model to perform similarly to a clean one. Our method demonstrates substantial advantages over several state-of-the-art methods, as evidenced by experiments on two datasets, two types of backdoor methods, and various attack configurations. Notably, our approach requires only a small portion of the training data -- only 1\% -- to purify the backdoor and reduce the attack success rate from 100\% to almost 0\%, a 100-fold improvement over the baseline methods. Our code is available at \url{https://github.com/judydnguyen/pbp-backdoor-purification-official}., Comment: Accepted at NDSS 2025

Uncertainty quantification in text-to-image (T2I) generative models is crucial for understanding model behavior and improving output reliability. In this paper, we are the first to quantify and evaluate the uncertainty of T2I models with respect to the prompt. Alongside adapting existing approaches designed to measure uncertainty in the image space, we also introduce Prompt-based UNCertainty Estimation for T2I models (PUNC), a novel method leveraging Large Vision-Language Models (LVLMs) to better address uncertainties arising from the semantics of the prompt and generated images. PUNC utilizes a LVLM to caption a generated image, and then compares the caption with the original prompt in the more semantically meaningful text space. PUNC also enables the disentanglement of both aleatoric and epistemic uncertainties via precision and recall, which image-space approaches are unable to do. Extensive experiments demonstrate that PUNC outperforms state-of-the-art uncertainty estimation techniques across various settings. Uncertainty quantification in text-to-image generation models can be used on various applications including bias detection, copyright protection, and OOD detection. We also introduce a comprehensive dataset of text prompts and generation pairs to foster further research in uncertainty quantification for generative models. Our findings illustrate that PUNC not only achieves competitive performance but also enables novel applications in evaluating and improving the trustworthiness of text-to-image models., Comment: 28 pages and 22 figures

Geodesic nets are types of graphs in Riemannian manifolds where each edge is a geodesic segment. One important object used in the construction of geodesic nets is a balanced vertex, where the sum of unit tangent vectors along adjacent edges is zero. In 2021, Parsch proved the upper bound for the number of balanced vertices of a geodesic net with three unbalanced vertices on surfaces with non-positive curvature. We extend his result by proving the existence of a balanced vertex of a triangle (with three unbalanced vertices) on any two-dimensional surface when all angles measure less than $2\pi/3$, if the length of the sides of the triangle are not too large. This property is also a generalization for the existence of the Fermat point of a planar triangle.

Contrastive learning has significantly improved representation quality, enhancing knowledge transfer across tasks in continual learning (CL). However, catastrophic forgetting remains a key challenge, as contrastive based methods primarily focus on "soft relationships" or "softness" between samples, which shift with changing data distributions and lead to representation overlap across tasks. Recently, the newly identified Neural Collapse phenomenon has shown promise in CL by focusing on "hard relationships" or "hardness" between samples and fixed prototypes. However, this approach overlooks "softness", crucial for capturing intra-class variability, and this rigid focus can also pull old class representations toward current ones, increasing forgetting. Building on these insights, we propose Focal Neural Collapse Contrastive (FNC^2), a novel representation learning loss that effectively balances both soft and hard relationships. Additionally, we introduce the Hardness-Softness Distillation (HSD) loss to progressively preserve the knowledge gained from these relationships across tasks. Our method outperforms state-of-the-art approaches, particularly in minimizing memory reliance. Remarkably, even without the use of memory, our approach rivals rehearsal-based methods, offering a compelling solution for data privacy concerns., Comment: Accepted at WACV 2025

3D Gaussian Splatting has advanced radiance field reconstruction, enabling high-quality view synthesis and fast rendering in 3D modeling. While adversarial attacks on object detection models are well-studied for 2D images, their impact on 3D models remains underexplored. This work introduces the Masked Iterative Fast Gradient Sign Method (M-IFGSM), designed to generate adversarial noise targeting the CLIP vision-language model. M-IFGSM specifically alters the object of interest by focusing perturbations on masked regions, degrading the performance of CLIP's zero-shot object detection capability when applied to 3D models. Using eight objects from the Common Objects 3D (CO3D) dataset, we demonstrate that our method effectively reduces the accuracy and confidence of the model, with adversarial noise being nearly imperceptible to human observers. The top-1 accuracy in original model renders drops from 95.4\% to 12.5\% for train images and from 91.2\% to 35.4\% for test images, with confidence levels reflecting this shift from true classification to misclassification, underscoring the risks of adversarial attacks on 3D models in applications such as autonomous driving, robotics, and surveillance. The significance of this research lies in its potential to expose vulnerabilities in modern 3D vision models, including radiance fields, prompting the development of more robust defenses and security measures in critical real-world applications., Comment: Accepted to Safe Generative AI Workshop @ NeurIPS 2024: https://neurips.cc/virtual/2024/workshop/84705

Diffusion models have shown remarkable abilities in generating realistic and high-quality images from text prompts. However, a trained model remains black-box; little do we know about the role of its components in exhibiting a concept such as objects or styles. Recent works employ causal tracing to localize layers storing knowledge in generative models without showing how those layers contribute to the target concept. In this work, we approach the model interpretability problem from a more general perspective and pose a question: \textit{``How do model components work jointly to demonstrate knowledge?''}. We adapt component attribution to decompose diffusion models, unveiling how a component contributes to a concept. Our framework allows effective model editing, in particular, we can erase a concept from diffusion models by removing positive components while remaining knowledge of other concepts. Surprisingly, we also show there exist components that contribute negatively to a concept, which has not been discovered in the knowledge localization approach. Experimental results confirm the role of positive and negative components pinpointed by our framework, depicting a complete view of interpreting generative models. Our code is available at \url{https://github.com/mail-research/CAD-attribution4diffusion}

Using the Cornwall-Jackiw-Tomboulis effective action framework in conjunction with variational perturbation theory, this study investigates the relative shift in the transition temperature of a homogeneous repulsive weakly interacting Bose gas compared to that of an ideal Bose gas. Employing both one-loop and self-consistent Popov approximations, we derive the universal form of the relative shift in the transition temperature, which is proportional to the s-wave scattering length. The results demonstrate excellent agreement with those obtained through precise Monte Carlo simulations. Furthermore, the zero-point energy and various thermodynamic quantities are also analyzed in the condensed phase.

The integration of deep learning tools in gastrointestinal vision holds the potential for significant advancements in diagnosis, treatment, and overall patient care. A major challenge, however, is these tools' tendency to make overconfident predictions, even when encountering unseen or newly emerging disease patterns, undermining their reliability. We address this critical issue of reliability by framing it as an out-of-distribution (OOD) detection problem, where previously unseen and emerging diseases are identified as OOD examples. However, gastrointestinal images pose a unique challenge due to the overlapping feature representations between in- Distribution (ID) and OOD examples. Existing approaches often overlook this characteristic, as they are primarily developed for natural image datasets, where feature distinctions are more apparent. Despite the overlap, we hypothesize that the features of an in-distribution example will cluster closer to the centroids of their ground truth class, resulting in a shorter distance to the nearest centroid. In contrast, OOD examples maintain an equal distance from all class centroids. Based on this observation, we propose a novel nearest-centroid distance deficit (NCCD) score in the feature space for gastrointestinal OOD detection. Evaluations across multiple deep learning architectures and two publicly available benchmarks, Kvasir2 and Gastrovision, demonstrate the effectiveness of our approach compared to several state-of-the-art methods. The code and implementation details are publicly available at: https://github.com/bhattarailab/NCDD

Multimodal Large Language Models (MLLMs) have become increasingly important due to their state-of-the-art performance and ability to integrate multiple data modalities, such as text, images, and audio, to perform complex tasks with high accuracy. This paper presents a comprehensive survey on personalized multimodal large language models, focusing on their architecture, training methods, and applications. We propose an intuitive taxonomy for categorizing the techniques used to personalize MLLMs to individual users, and discuss the techniques accordingly. Furthermore, we discuss how such techniques can be combined or adapted when appropriate, highlighting their advantages and underlying rationale. We also provide a succinct summary of personalization tasks investigated in existing research, along with the evaluation metrics commonly used. Additionally, we summarize the datasets that are useful for benchmarking personalized MLLMs. Finally, we outline critical open challenges. This survey aims to serve as a valuable resource for researchers and practitioners seeking to understand and advance the development of personalized multimodal large language models.

Topological correctness is critical for segmentation of tubular structures. Existing topological segmentation loss functions are primarily based on the persistent homology of the image. They match the persistent features from the segmentation with the persistent features from the ground truth and minimize the difference between them. However, these methods suffer from an ambiguous matching problem since the matching only relies on the information in the topological space. In this work, we propose an effective and efficient Spatial-Aware Topological Loss Function that further leverages the information in the original spatial domain of the image to assist the matching of persistent features. Extensive experiments on images of various types of tubular structures show that the proposed method has superior performance in improving the topological accuracy of the segmentation compared with state-of-the-art methods.

Handling occlusion remains a significant challenge for video instance-level tasks like Multiple Object Tracking (MOT) and Video Instance Segmentation (VIS). In this paper, we propose a novel framework, Amodal-Aware Video Instance Segmentation (A2VIS), which incorporates amodal representations to achieve a reliable and comprehensive understanding of both visible and occluded parts of objects in a video. The key intuition is that awareness of amodal segmentation through spatiotemporal dimension enables a stable stream of object information. In scenarios where objects are partially or completely hidden from view, amodal segmentation offers more consistency and less dramatic changes along the temporal axis compared to visible segmentation. Hence, both amodal and visible information from all clips can be integrated into one global instance prototype. To effectively address the challenge of video amodal segmentation, we introduce the spatiotemporal-prior Amodal Mask Head, which leverages visible information intra clips while extracting amodal characteristics inter clips. Through extensive experiments and ablation studies, we show that A2VIS excels in both MOT and VIS tasks in identifying and tracking object instances with a keen understanding of their full shape., Comment: Preprint. Project page: https://uark-aicv.github.io/A2VIS

ALMA-IMF imaged 15 massive protoclusters down to a resolution of of 2 kau scales, identifying about 1000 star-forming cores. The mass and luminosity of these cores, which are fundamental physical characteristics, are difficult to determine, a problem greatly exacerbated at the distances >2 kpc of ALMA-IMF protoclusters. We combined new datasets and radiative transfer modeling to characterize these cores. We estimated their mass-averaged temperature and the masses these estimates imply. For 1/6 of the sample, we measured the bolometric luminosities, implementing deblending corrections when necessary. We used spectral energy distribution (SED) analysis obtained with the PPMAP Bayesian procedure, which aims to preserve the best angular resolution of the input data. We extrapolated the luminosity and dust temperature images provided by PPMAP at 2.5" resolution to estimate those of individual cores, which were identified at higher angular resolution. To do this, we applied approximate radiative transfer relationships between the luminosity of a protostar and the temperature of its surrounding envelope and between the external heating of prestellar cores and their temperatures. For the first time, we provide data-informed estimates of dust temperatures for 883 cores identified with ALMA-IMF: 17-31 K and 28-79 K (5th and 95th percentiles, up to 127 K) for the 617 prestellar and 266 protostellar cores, respectively. We also measured protostellar luminosities spanning 20-80 000 Lsun. For hot cores, we estimated systematically lower temperatures than studies based on complex organic molecules. We established a mass-luminosity evolutionary diagram, for the first time at the core spatial resolution and for a large sample of high-mass protostellar cores. The ALMA-IMF data favor a scenario in which protostars accrete their mass from a larger mass reservoir than their host cores., Comment: 42 pages, 13 figures, accepted by A&A

Efficiently modeling large 2D contexts is essential for various fields including Giga-Pixel Whole Slide Imaging (WSI) and remote sensing. Transformer-based models offer high parallelism but face challenges due to their quadratic complexity for handling long sequences. Recently, Mamba introduced a selective State Space Model (SSM) with linear complexity and high parallelism, enabling effective and efficient modeling of wide context in 1D sequences. However, extending Mamba to vision tasks, which inherently involve 2D structures, results in spatial discrepancies due to the limitations of 1D sequence processing. On the other hand, current 2D SSMs inherently model 2D structures but they suffer from prohibitively slow computation due to the lack of efficient parallel algorithms. In this work, we propose 2DMamba, a novel 2D selective SSM framework that incorporates the 2D spatial structure of images into Mamba, with a highly optimized hardware-aware operator, adopting both spatial continuity and computational efficiency. We validate the versatility of our approach on both WSIs and natural images. Extensive experiments on 10 public datasets for WSI classification and survival analysis show that 2DMamba~improves up to $2.48\%$ in AUC, $3.11\%$ in F1 score, $2.47\%$ in accuracy and $5.52\%$ in C-index. Additionally, integrating our method with VMamba for natural imaging yields $0.5$ to $0.7$ improvements in mIoU on the ADE20k semantic segmentation dataset, and $0.2\%$ accuracy improvement on ImageNet-1K classification dataset. Our code is available at https://github.com/AtlasAnalyticsLab/2DMamba., Comment: Submission under review

The asymmetric coloring number of a graph is the minimum number of colors needed to color its vertices, so that no non-trivial automorphism preserves the color classes. We investigate the asymmetric coloring number of graphs that are disjoint unions of graphs. We will derive a general relationship between the asymmetric coloring number of disjoint copies of graphs and the number of ways to color a single copy asymmetrically, and then look at particular cases such as disjoint copies of paths, stars, cycles, and hypercubes., Comment: Lack of Inclusion/Approval of Co-Authors

Photoreforming is a sustainable photocatalytic process that degrades plastic waste while simultaneously producing hydrogen (H2) from water. However, this process has received limited attention due to the scarcity of effective catalysts capable of both plastic degradation and H2 production, such as titanium dioxide (TiO2). In this study, an active catalyst is developed by stabilizing the high-pressure orthorhombic phase of TiO2, known as columbite, using a high pressure torsion (HPT) method. This material effectively degrades polyethylene terephthalate (PET) plastic under light, converting it into valuable organic compounds such as formic acid, terephthalate, glycolic acid, and acetic acid. Additionally, it produces a significant amount of H2. The findings show that the high-pressure orthorhombic phase, especially in the presence of oxygen vacancies, enhances catalytic H2 production and microplastic degradation by increasing light absorption, reducing electron-hole recombination, and generating hydroxyl radicals. These results highlight the substantial potential of modified high-pressure TiO2 photocatalysts in simultaneously addressing the plastic waste crisis and the demand for H2 fuel.

Bismuth (III) oxide (Bi2O3) has been highly studied as a photocatalyst for green hydrogen production due to its low band gap, yet its efficiency requires enhancement. This study synthesizes a defective and strained black Bi2O3 by severe straining under high pressure, via a high-pressure torsion method, to improve its photocatalytic hydrogen production. The material rich in oxygen vacancies exhibits a ten-fold improvement in water splitting with excellent cycling stability. Such improvement is due to improved light absorption, narrowing band gap and reduced irradiative electron-hole recombination. Moreover, the valence band bottom energy positively increases by straining leading to a high overpotential for hydrogen production. This research highlights the potential of vacancies and lattice strain in developing dopant-free active catalysts for water splitting.

Simultaneous catalytic hydrogen (H2) production and plastic waste degradation under light, known as photoreforming, is a novel approach to green fuel production and efficient waste management. Here, we use a high-entropy oxide (HEO), a new family of catalysts with five or more principal cations in their structure, for plastic degradation and simultaneous H2 production. The HEO shows higher activity than that of P25 TiO2, a benchmark photocatalyst, for the degradation of polyethylene terephthalate (PET) plastics in water. Several valuable products are produced by photoreforming of PET bottles and microplastics including H2, terephthalate, ethylene glycol and formic acid. The high activity is attributed to the diverse existence of several cations in the HEO lattice, lattice defects, and appropriate charge carrier lifetime. These findings suggest that HEOs possess high potential as new catalysts for concurrent plastic waste conversion and clean H2 production.

We propose an electrolyser model for steady-state load flow analysis of multi-carrier energy networks, where the electrolyser is capable of producing hydrogen gas and heat. We show that there are boundary conditions that lead to a well-posed problem. We derive these conditions for two cases, namely with a fixed and non-fixed ratio between gas and heat output. Furthermore, the derived conditions are validated numerically.

Traditional deep learning methods in medical imaging often focus solely on segmentation or classification, limiting their ability to leverage shared information. Multi-task learning (MTL) addresses this by combining both tasks through shared representations but often struggles to balance local spatial features for segmentation and global semantic features for classification, leading to suboptimal performance. In this paper, we propose a simple yet effective UNet-based MTL model, where features extracted by the encoder are used to predict classification labels, while the decoder produces the segmentation mask. The model introduces an advanced encoder incorporating a novel ResFormer block that integrates local context from convolutional feature extraction with long-range dependencies modeled by the Transformer. This design captures broader contextual relationships and fine-grained details, improving classification and segmentation accuracy. To enhance classification performance, multi-scale features from different encoder levels are combined to leverage the hierarchical representation of the input image. For segmentation, the features passed to the decoder via skip connections are refined using a novel dilated feature enhancement (DFE) module, which captures information at different scales through three parallel convolution branches with varying dilation rates. This allows the decoder to detect lesions of varying sizes with greater accuracy. Experimental results across multiple medical datasets confirm the superior performance of our model in both segmentation and classification tasks, compared to state-of-the-art single-task and multi-task learning methods.

Semiparametric mixture models are parametric models with latent variables. They are defined kernel, $p_\theta(x | z)$, where z is the unknown latent variable, and $\theta$ is the parameter of interest. We assume that the latent variables are an i.i.d. sample from some mixing distribution $F$. A Bayesian would put a prior on the pair $(\theta, F)$. We prove consistency for these models in fair generality and then study efficiency. We first prove an abstract Semiparametric Bernstein-von Mises theorem, and then provide tools to verify the assumptions. We use these tools to study the efficiency for estimating $\theta$ in the frailty model and the errors in variables model in the case were we put a generic prior on $\theta$ and a species sampling process prior on $F$.

Industrial networks are undergoing rapid transformation driven by the convergence of emerging technologies that are revolutionizing conventional workflows, enhancing operational efficiency, and fundamentally redefining the industrial landscape across diverse sectors. Amidst this revolution, Digital Twin (DT) emerges as a transformative innovation that seamlessly integrates real-world systems with their virtual counterparts, bridging the physical and digital realms. In this article, we present a comprehensive survey of the emerging DT-enabled services and applications across industries, beginning with an overview of DT fundamentals and its components to a discussion of key enabling technologies for DT. Different from literature works, we investigate and analyze the capabilities of DT across a wide range of industrial services, including data sharing, data offloading, integrated sensing and communication, content caching, resource allocation, wireless networking, and metaverse. In particular, we present an in-depth technical discussion of the roles of DT in industrial applications across various domains, including manufacturing, healthcare, transportation, energy, agriculture, space, oil and gas, as well as robotics. Throughout the technical analysis, we delve into real-time data communications between physical and virtual platforms to enable industrial DT networking. Subsequently, we extensively explore and analyze a wide range of major privacy and security issues in DT-based industry. Taxonomy tables and the key research findings from the survey are also given, emphasizing important insights into the significance of DT in industries. Finally, we point out future research directions to spur further research in this promising area.

We investigate the decay of bosonic dark matter with masses between 1 MeV and 2 TeV into Standard Model final states. We specifically focus on dark photons that kinetically mix with the Standard Model, as well as scalar dark matter models that have Yukawa couplings with the Standard Model. Using INTEGRAL and AMS-02 data, we constrain the dark matter decay lifetime into final states that include photons or positrons, setting strong constraints on the dark matter lifetime that reach 10$^{25}$ s for dark matter below 10 GeV and up to 10$^{29}$ s for dark matter above 10 GeV., Comment: 13 pages, 8 figures. Appendix: 2 pages, 1 figure. Comments are welcome!

Traction force microscopy is a method widely used in biophysics and cell biology to determine forces that biological cells apply to their environment. In the experiment, the cells adhere to a soft elastic substrate, which is then deformed in response to cellular traction forces. The inverse problem consists in computing the traction stress applied by the cell from microscopy measurements of the substrate deformations. In this work, we consider a linear model, in which 3D forces are applied at a 2D interface, called 2.5D traction force microscopy, and a nonlinear pure 2D model, from which we directly obtain a linear pure 2D model. All models lead to a linear resp. nonlinear parameter identification problem for a boundary value problem of elasticity. We analyze the respective forward operators and conclude with some numerical experiments for simulated and experimental data., Comment: 28 pages, 9 figures

This paper addresses the state estimation problem in continuous LTI systems under attacks with non-periodic and asynchronous sampled measurements. The non-periodic and asynchronous sampling requires sensors to transmit not only the measurement values but also the sampling time-stamps to the fusion center via unprotected communication channels. This communication scheme leaves the system vulnerable to a variety of malicious activities such as (i) manipulating measurement values, (ii) manipulating time-stamps, (iii) hybrid manipulations such as generating fake measurements or eliminating the measurement. To deal with such more powerful attacks, we propose a decentralized local estimation algorithm where each sensor maintains its local state estimate based on its measurements in an asynchronous fashion. The local states are synchronized by time-prediction and fused in an event-triggered manner. In the absence of attacks, local estimates are proved to recover the optimal Kalman estimation by our carefully designed weighted least square problem, given that the sample time is non-pathological. In the presence of attacks, an $\ell_1$ regularized least square problem is proposed to generate secure estimates with uniformly bounded error as long as the observability redundancy is satisfied. The effectiveness of the proposed algorithm is demonstrated through a benchmark example of the IEEE 14-bus system., Comment: 9 pages and 6 figures. arXiv admin note: text overlap with arXiv:2303.17514

Social media platforms play a pivotal role in shaping public opinion and amplifying political discourse, particularly during elections. However, the same dynamics that foster democratic engagement can also exacerbate polarization. To better understand these challenges, here, we investigate the ideological positioning of tweets related to the 2024 U.S. Presidential Election. To this end, we analyze 1,235 tweets from key political figures and 63,322 replies, and classify ideological stances into Pro-Democrat, Anti-Republican, Pro-Republican, Anti-Democrat, and Neutral categories. Using a classification pipeline involving three large language models (LLMs)-GPT-4o, Gemini-Pro, and Claude-Opus-and validated by human annotators, we explore how ideological alignment varies between candidates and constituents. We find that Republican candidates author significantly more tweets in criticism of the Democratic party and its candidates than vice versa, but this relationship does not hold for replies to candidate tweets. Furthermore, we highlight shifts in public discourse observed during key political events. By shedding light on the ideological dynamics of online political interactions, these results provide insights for policymakers and platforms seeking to address polarization and foster healthier political dialogue., Comment: 5 pages, 2 figures

Let $f: {\mathbb Z}_N^d \to {\mathbb C}$ be a signal with the Fourier transform $\widehat{f}: \Bbb Z_N^d\to \Bbb C$. A classical result due to Matolcsi and Szucs (\cite{MS73}), and, independently, to Donoho and Stark (\cite{DS89}) states if a subset of frequencies ${\{\widehat{f}(m)\}}_{m \in S}$ of $f$ are unobserved due to noise or other interference, then $f$ can be recovered exactly and uniquely provided that $$ |E| \cdot |S|<\frac{N^d}{2},$$ where $E$ is the support of $f$, i.e., $E=\{x \in {\mathbb Z}_N^d: f(x) \not=0\}$. In this paper, we consider signals that are Dirac combs of complexity $\gamma$, meaning they have the form $f(x)=\sum_{i=1}^{\gamma} a_i 1_{A_i}(x)$, where the sets $A_i \subset {\mathbb Z}_N^d$ are disjoint, $a_i$ are complex numbers, and $\gamma \leq N^d$. We will define the concept of effective support of these signals and show that if $\gamma$ is not too large, a good recovery condition can be obtained by pigeonholing under additional reasonable assumptions on the distribution of values. Our approach produces a non-trivial uncertainty principle and a signal recovery condition in many situations when the support of the function is too large to apply the classical theory.

Clinical language models are important for many applications in healthcare, but their development depends on access to extensive clinical text for pretraining. However, obtaining clinical notes from electronic health records (EHRs) at scale is challenging due to patient privacy concerns. In this study, we rephrase existing clinical notes using LLMs to generate synthetic pretraining corpora, drawing inspiration from previous work on rephrasing web data. We examine four popular small-sized LLMs (<10B) to create synthetic clinical text to pretrain both decoder-based and encoder-based language models. The method yields better results in language modeling and downstream tasks than previous synthesis approaches without referencing real clinical text. We find that augmenting original clinical notes with synthetic corpora from different LLMs improves performances even at a small token budget, showing the potential of this method to support pretraining at the institutional level or be scaled to synthesize large-scale clinical corpora.

Recent years have seen soft robotic grippers gain increasing attention due to their ability to robustly grasp soft and fragile objects. However, a commonly available standardised evaluation protocol has not yet been developed to assess the performance of varying soft robotic gripper designs. This work introduces a novel protocol, the Soft Grasping Benchmarking and Evaluation (SoGraB) method, to evaluate grasping quality, which quantifies object deformation by using the Density-Aware Chamfer Distance (DCD) between point clouds of soft objects before and after grasping. We validated our protocol in extensive experiments, which involved ranking three Fin-Ray gripper designs with a subset of the EGAD object dataset. The protocol appropriately ranked grippers based on object deformation information, validating the method's ability to select soft grippers for complex grasping tasks and benchmark them for comparison against future designs., Comment: 6 pages, 7 figures

Adapting vision transformer foundation models through parameter-efficient fine-tuning (PEFT) methods has become increasingly popular. These methods optimize a limited subset of parameters, enabling efficient adaptation without the need to fine-tune the entire model while still achieving competitive performance. However, traditional PEFT methods may limit the model's capacity to capture complex patterns, especially those associated with high-frequency spectra. This limitation becomes particularly problematic as existing research indicates that high-frequency features are crucial for distinguishing subtle image structures. To address this issue, we introduce FreqFit, a novel Frequency Fine-tuning module between ViT blocks to enhance model adaptability. FreqFit is simple yet surprisingly effective, and can be integrated with all existing PEFT methods to boost their performance. By manipulating features in the frequency domain, our approach allows models to capture subtle patterns more effectively. Extensive experiments on 24 datasets, using both supervised and self-supervised foundational models with various state-of-the-art PEFT methods, reveal that FreqFit consistently improves performance over the original PEFT methods with performance gains ranging from 1% to 16%. For instance, FreqFit-LoRA surpasses the performances of state-of-the-art baselines on CIFAR100 by more than 10% even without applying regularization or strong augmentation. For reproducibility purposes, the source code is available at https://github.com/tsly123/FreqFiT., Comment: 24 pages

The problem of minimizing the difference of two lower semicontinuous, proper, convex functions (a DC function) on a nonempty closed convex set in a locally convex Hausdorff topological vector space is studied in this paper. The focus is made on the situations where either the second component of the objective function is a generalized polyhedral convex function or the first component of the objective function is a generalized polyhedral convex function and the constraint set is generalized polyhedral convex. Various results on optimality conditions, the local solution set, the global solution set, and solution algorithms via duality are obtained. Useful illustrative examples are considered.

We provide a detailed decomposition of Wigner's particles, defined as unitary irreducible representations of the Poincar\'e group, in terms of unitary representations of its Lorentz subgroup. As pointed out before us, this decomposition only involves Lorentz representations belonging to the principal continuous series, and further underpins the connection between scattering amplitudes and conformal correlation functions discussed in the context of celestial holography. We provide very explicit formulae for the decomposition of particles of arbitrary mass and (half-)integer spin and for any spacetime dimension. We emphasise that this decomposition is unique and comes with a specific inner product on the corresponding Hilbert space. We also confirm that unitary translations mix Lorentz representations within the principal continuous series only, as required. Implications to the celestial holography program are indicated., Comment: 33 pages

Transformers exhibit in-context learning (ICL): the ability to use novel information presented in the context without additional weight updates. Recent work shows that ICL emerges when models are trained on a sufficiently diverse set of tasks and the transition from memorization to generalization is sharp with increasing task diversity. One interpretation is that a network's limited capacity to memorize favors generalization. Here, we examine the mechanistic underpinnings of this transition using a small transformer applied to a synthetic ICL task. Using theory and experiment, we show that the sub-circuits that memorize and generalize can be viewed as largely independent. The relative rates at which these sub-circuits learn explains the transition from memorization to generalization, rather than capacity constraints. We uncover a memorization scaling law, which determines the task diversity threshold at which the network generalizes. The theory quantitatively explains a variety of other ICL-related phenomena, including the long-tailed distribution of when ICL is acquired, the bimodal behavior of solutions close to the task diversity threshold, the influence of contextual and data distributional statistics on ICL, and the transient nature of ICL.

Speaker-attributed automatic speech recognition (SA-ASR) aims to transcribe speech while assigning transcripts to the corresponding speakers accurately. Existing methods often rely on complex modular systems or require extensive fine-tuning of joint modules, limiting their adaptability and general efficiency. This paper introduces a novel approach, leveraging a frozen multilingual ASR model to incorporate speaker attribution into the transcriptions, using only standard monolingual ASR datasets. Our method involves training a speaker module to predict speaker embeddings based on weak labels without requiring additional ASR model modifications. Despite being trained exclusively with non-overlapping monolingual data, our approach effectively extracts speaker attributes across diverse multilingual datasets, including those with overlapping speech. Experimental results demonstrate competitive performance compared to strong baselines, highlighting the model's robustness and potential for practical applications.

Convolutional Neural Networks (CNNs) achieve high performance in image classification tasks but are challenging to deploy on resource-limited hardware due to their large model sizes. To address this issue, we leverage Mutual Information, a metric that provides valuable insights into how deep learning models retain and process information through measuring the shared information between input features or output labels and network layers. In this study, we propose a structured filter-pruning approach for CNNs that identifies and selectively retains the most informative features in each layer. Our approach successively evaluates each layer by ranking the importance of its feature maps based on Conditional Mutual Information (CMI) values, computed using a matrix-based Renyi {\alpha}-order entropy numerical method. We propose several formulations of CMI to capture correlation among features across different layers. We then develop various strategies to determine the cutoff point for CMI values to prune unimportant features. This approach allows parallel pruning in both forward and backward directions and significantly reduces model size while preserving accuracy. Tested on the VGG16 architecture with the CIFAR-10 dataset, the proposed method reduces the number of filters by more than a third, with only a 0.32% drop in test accuracy.

This paper presents a preliminary study of an efficient object tracking approach, comparing the performance of two different 3D point cloud sensory sources: LiDAR and stereo cameras, which have significant price differences. In this preliminary work, we focus on single object tracking. We first developed a fast heuristic object detector that utilizes prior information about the environment and target. The resulting target points are subsequently fed into an extended object tracking framework, where the target shape is parameterized using a star-convex hypersurface model. Experimental results show that our object tracking method using a stereo camera achieves performance similar to that of a LiDAR sensor, with a cost difference of more than tenfold.

This paper investigates nonlinear Landau damping in the 3D Vlasov-Poisson (VP) system. We study the asymptotic stability of the Poisson equilibrium $\mu(v)=\frac{1}{\pi^2(1+|v|^2)^2}$ under small perturbations. Building on the foundational work of Ionescu, Pausader, Wang, and Widmayer \cite{AIonescu2022}, we provide a streamlined proof of nonlinear Landau damping for the 3D unscreened VP system. Our analysis leverages sharp decay estimates, novel decomposition techniques to demonstrate the stabilization of the particle distribution and the decay of electric field. These results reveal the free transport-like behavior for the perturbed density $\rho(t,x)$, and enhance the understanding of Landau damping in an unconfined setting near stable equilibria., Comment: 13 pages

The rapid integration of Artificial Intelligence (AI)-based systems offers benefits for various domains of the economy and society but simultaneously raises concerns due to emerging scandals. These scandals have led to the increasing importance of AI accountability to ensure that actors provide justification and victims receive compensation. However, AI accountability has a negative connotation due to its emphasis on penalizing sanctions, resulting in reactive approaches to emerging concerns. To counteract the prevalent negative view and offer a proactive approach to facilitate the AI accountability behavior of developers, we explore rewards as an alternative mechanism to sanctions. We develop a theoretical model grounded in Self-Determination Theory to uncover the potential impact of rewards and sanctions on AI developers. We further identify typical sanctions and bug bounties as potential reward mechanisms by surveying related research from various domains, including cybersecurity., Comment: Accepted for presentation at the 34th Workshop on Information Technologies and Systems (WITS)

Iron oxide (e.g., Fe$_3$O$_4$ or Fe$_2$O$_3$) nanoparticles are promising candidates for a variety of biomedical applications ranging from magnetic hyperthermia therapy to drug delivery and bio-detection, due to their superparamagnetism, non-toxicity, and biodegradability. While particles of small size (below a critical size, ~20 nm) display superparamagnetic behavior at room temperature, these particles tend to penetrate highly sensitive areas of the body such as the Blood-Brain Barrier (BBB), leading to undesired effects. In addition, these particles possess a high probability of retention, which can lead to genotoxicity and biochemical toxicity. Increasing particle size is a means for addressing these problems but also suppresses the superparamagnetism. We have overcome this particle size limit by synthesizing unique polycrystalline iron oxide nanoparticles composed of multiple nanocrystals of 10 to 15 nm size while tuning particle size from 160 to 400 nm. These so-called superparticles preserve superparamagnetic characteristics and exhibit excellent hyperthermia responses. The specific absorption rates (SAR) exceed 250 W/g (HAC = 800 Oe, f = 310 kHz) at a low concentration of 0.5 mg/mL, indicating their capability in cancer treatment with minimum dose. Our study underscores the potential of size-tunable polycrystalline iron oxide superparticles with superparamagnetic properties for advanced biomedical applications and sensing technologies.

Typical deep neural video compression networks usually follow the hybrid approach of classical video coding that contains two separate modules: motion coding and residual coding. In addition, a symmetric auto-encoder is often used as a normal architecture for both motion and residual coding. In this paper, we propose a novel approach that handles the drawbacks of the two typical above-mentioned architectures, we call it kernel-based motion-free video coding. The advantages of the motion-free approach are twofold: it improves the coding efficiency of the network and significantly reduces computational complexity thanks to eliminating motion estimation, motion compensation, and motion coding which are the most time-consuming engines. In addition, the kernel-based auto-encoder alleviates blur artifacts that usually occur with the conventional symmetric autoencoder. Consequently, it improves the visual quality of the reconstructed frames. Experimental results show the proposed framework outperforms the SOTA deep neural video compression networks on the HEVC-class B dataset and is competitive on the UVG and MCL-JCV datasets. In addition, it generates high-quality reconstructed frames in comparison with conventional motion coding-based symmetric auto-encoder meanwhile its model size is much smaller than that of the motion-based networks around three to four times., Comment: Deep Neural Video Compression

Noisy labels pose a substantial challenge in machine learning, often resulting in overfitting and poor generalization. Sharpness-Aware Minimization (SAM), as demonstrated in Foret et al. (2021), improves generalization over traditional Stochastic Gradient Descent (SGD) in classification tasks with noisy labels by implicitly slowing noisy learning. While SAM's ability to generalize in noisy environments has been studied in several simplified settings, its full potential in more realistic training settings remains underexplored. In this work, we analyze SAM's behavior at each iteration, identifying specific components of the gradient vector that contribute significantly to its robustness against noisy labels. Based on these insights, we propose SANER (Sharpness-Aware Noise-Explicit Reweighting), an effective variant that enhances SAM's ability to manage noisy fitting rate. Our experiments on CIFAR-10, CIFAR-100, and Mini-WebVision demonstrate that SANER consistently outperforms SAM, achieving up to an 8% increase on CIFAR-100 with 50% label noise.

A pulsar's scintillation bandwidth is inversely proportional to the scattering delay, making accurate measurements of scintillation bandwidth critical to characterize unmitigated delays in efforts to measure low-frequency gravitational waves with pulsar timing arrays. In this pilot work, we searched for a subset of known pulsars within $\sim$97% of the data taken with the PUPPI instrument for the AO327 survey with the Arecibo telescope, attempting to measure the scintillation bandwidths in the dataset by fitting to the 2D autocorrelation function of their dynamic spectra. We successfully measured 38 bandwidths from 23 pulsars (six without prior literature values), finding that: almost all of the measurements are larger than the predictions from NE2001 and YMW16 (two popular galactic models); NE2001 is more consistent with our measurements than YMW16; Gaussian fits to the bandwidth are more consistent with both electron density models than Lorentzian ones; and for the 17 pulsars with prior literature values, the measurements between various sources often vary by factors of a few. The success of Gaussian fits may be due to the use of Gaussian fits to train models in previous work. The variance of literature values over time could relate to the scaling factor used to compare measurements, but also seems consistent with time-varying interstellar medium parameters. This work can be extended to the rest of AO327 to further investigate these trends, highlighting the continuing importance of large archival datasets for projects beyond their initial conception., Comment: 21 pages, 8 figures, 2 tables

Air quality has important climate and health effects. There is a need, therefore, to monitor air quality both indoors and outdoors. Methods of measuring air quality should be cost-effective if they are to be used widely, and one such method is low-cost sensors (LCS). This study reports on the use of LCSs in Ulaanbaatar, Mongolia, to measure $\mathrm{PM_{2.5}}$ concentrations inside yurts or "gers." Some of these gers were part of a non-government agency (NGO) initiative to improve the insulating properties of these housing structures. The goal of the NGO was to decrease particulate emissions inside the gers; a secondary result was to lower the use of coal and other biomass material. LCSs were installed in gers heated primarily by coal, and interior air quality was measured. Gers that were modified by increasing their insulating capacities showed a 17.5% reduction in $\mathrm{PM_{2.5}}$ concentrations, but these concentrations remained higher than levels recommended by health organizations. Gers that were insulated and used a combination of both coal and electricity showed a 19.1% reduction in $\mathrm{PM_{2.5}}$ concentrations. Insulated gers that used electricity for both heating and cooking showed a 48% reduction in $\mathrm{PM_{2.5}}$, though concentrations were still 6.4 times higher than those recommended by the World Health Organization (WHO). Nighttime and daytime trends followed similar patterns in $\mathrm{PM_{2.5}}$ concentrations with slight variations. It was found that, at nighttime, the outside $\mathrm{PM_{2.5}}$ concentrations were generally higher than the inside concentrations of the gers in this study. This suggests that $\mathrm{PM_{2.5}}$ would flow into the gers whenever the doors were opened, causing spikes in $\mathrm{PM_{2.5}}$ concentrations.

Quantifying forest carbon is crucial for informing decisions and policies that will protect the planet. Machine learning (ML) and remote sensing (RS) techniques have been used to do this task more effectively, yet there lacks a systematic review on the most recent ML methods and RS combinations, especially with the consideration of forest characteristics. This study systematically analyzed 25 papers meeting strict inclusion criteria from over 80 related studies, identifying 28 ML methods and key combinations of RS data. Random Forest had the most frequent appearance (88\% of studies), while Extreme Gradient Boosting showed superior performance in 75\% of the studies in which it was compared with other methods. Sentinel-1 emerged as the most utilized remote sensing source, with multi-sensor approaches (e.g., Sentinel-1, Sentinel-2, and LiDAR) proving especially effective. Our findings provide grounds for recommending best practices in integrating machine learning and remote sensing for accurate and scalable forest carbon stock estimation., Comment: First author and corresponding author: Autumn Nguyen

Modern AI (i.e., Deep Learning and its variants) is here to stay. However, its enigmatic black box nature presents a fundamental challenge to the traditional methods of test and validation (T&E). Or does it? In this paper we introduce a Digital Engineering (DE) approach to T&E (DE-T&E), combined with generative AI, that can achieve requisite mil spec statistical validation as well as uncover potential deleterious Black Swan events that might otherwise not be uncovered until it is too late. An illustration of these concepts is presented for an advanced modern radar example employing deep learning AI.