Search

The proliferation of smartphones and other mobile devices provides a unique opportunity to make Advanced Driver Assistance Systems (ADAS) accessible to everyone in the form of an application empowered by low-cost Machine/Deep Learning (ML/DL) models to enhance road safety. For the critical feature of Collision Avoidance in Mobile ADAS, lightweight Deep Neural Networks (DNN) for object detection exist, but conventional pixel-wise depth/distance estimation DNNs are vastly more computationally expensive making them unsuitable for a real-time application on resource-constrained devices. In this paper, we present a distance estimation model, DECADE, that processes each detector output instead of constructing pixel-wise depth/disparity maps. In it, we propose a pose estimation DNN to estimate allocentric orientation of detections to supplement the distance estimation DNN in its prediction of distance using bounding box features. We demonstrate that these modules can be attached to any detector to extend object detection with fast distance estimation. Evaluation of the proposed modules with attachment to and fine-tuning on the outputs of the YOLO object detector on the KITTI 3D Object Detection dataset achieves state-of-the-art performance with 1.38 meters in Mean Absolute Error and 7.3% in Mean Relative Error in the distance range of 0-150 meters. Our extensive evaluation scheme not only evaluates class-wise performance, but also evaluates range-wise accuracy especially in the critical range of 0-70m., Comment: 8 pages, 17 figures, 4 tables

As Systems of Systems evolve into increasingly complex networks, harnessing their collective potential becomes paramount. Traditional SoS engineering approaches lack the necessary programmability to develop third party SoS level behaviors. To address this challenge, we propose a software defined approach to enable flexible and adaptive programming of SoS. We introduce the Holon Programming Model, a software-defined framework designed to meet these needs. The Holon Programming Model empowers developers to design and orchestrate complex system behaviors effectively, as illustrated in our disaster management scenario. This research outlines the Holon Programming Model theoretical underpinnings and practical applications, with the aim of driving further exploration and advancement in the field of software defined SoS

Pedestrian motion prediction is a key part of the modular-based autonomous driving pipeline, ensuring safe, accurate, and timely awareness of human agents' possible future trajectories. The autonomous vehicle can use this information to prevent any possible accidents and create a comfortable and pleasant driving experience for the passengers and pedestrians. A wealth of research was done on the topic from the authors of robotics, computer vision, intelligent transportation systems, and other fields. However, a relatively unexplored angle is the integration of the state-of-art solutions into existing autonomous driving stacks and evaluating them in real-life conditions rather than sanitized datasets. We analyze selected publications with provided open-source solutions and provide a perspective obtained by integrating them into existing Autonomous Driving framework - Autoware Mini and performing experiments in natural urban conditions in Tartu, Estonia to determine valuability of traditional motion prediction metrics. This perspective should be valuable to any potential autonomous driving or robotics engineer looking for the real-world performance of the existing state-of-art pedestrian motion prediction problem. The code with instructions on accessing the dataset is available at https://github.com/dmytrozabolotnii/autoware_mini., Comment: 7 pages, 2 figures, 4 tables This work has been submitted to the IEEE for possible publication

The linear canonical transform (LCT) serves as a powerful generalization of the Fourier transform (FT), encapsulating various integral transforms within a unified framework. This versatility has made it a cornerstone in fields such as signal processing, optics, and quantum mechanics. Extending this concept to quaternion algebra, the Quaternion Fourier Transform (QFT) emerged, enriching the analysis of multidimensional and complex-valued signals. The Quaternion Linear Canonical Transform (QLCT), a further generalization, has now positioned itself as a central tool across various disciplines, including applied mathematics, engineering, computer science, and statistics. In this paper, we introduce the Two Dimensional Quaternion Linear Canonical Transform (2DQLCT) as a novel framework for probability modeling. By leveraging the 2DQLCT, we aim to provide a more comprehensive understanding of probability distributions, particularly in the context of multi-dimensional and complex-valued signals. This framework not only broadens the theoretical underpinnings of probability theory but also opens new avenues for researchers

In this study, we propose a novel and robust framework, Self-DenseMobileNet, designed to enhance the classification of nodules and non-nodules in chest radiographs (CXRs). Our approach integrates advanced image standardization and enhancement techniques to optimize the input quality, thereby improving classification accuracy. To enhance predictive accuracy and leverage the strengths of multiple models, the prediction probabilities from Self-DenseMobileNet were transformed into tabular data and used to train eight classical machine learning (ML) models; the top three performers were then combined via a stacking algorithm, creating a robust meta-classifier that integrates their collective insights for superior classification performance. To enhance the interpretability of our results, we employed class activation mapping (CAM) to visualize the decision-making process of the best-performing model. Our proposed framework demonstrated remarkable performance on internal validation data, achieving an accuracy of 99.28\% using a Meta-Random Forest Classifier. When tested on an external dataset, the framework maintained strong generalizability with an accuracy of 89.40\%. These results highlight a significant improvement in the classification of CXRs with lung nodules., Comment: 31 pages

Single-cell RNA sequencing (scRNA-seq) enables the study of cellular diversity at single cell level. It provides a global view of cell-type specification during the onset of biological mechanisms such as developmental processes and human organogenesis. Various statistical, machine and deep learning-based methods have been proposed for cell-type classification. Most of the methods utilizes unsupervised lower dimensional projections obtained from for a large reference data. In this work, we proposed a reference-based method for cell type classification, called EnProCell. The EnProCell, first, computes lower dimensional projections that capture both the high variance and class separability through an ensemble of principle component analysis and multiple discriminant analysis. In the second phase, EnProCell trains a deep neural network on the lower dimensional representation of data to classify cell types. The proposed method outperformed the existing state-of-the-art methods when tested on four different data sets produced from different single-cell sequencing technologies. The EnProCell showed higher accuracy (98.91) and F1 score (98.64) than other methods for predicting reference from reference datasets. Similarly, EnProCell also showed better performance than existing methods in predicting cell types for data with unknown cell types (query) from reference datasets (accuracy:99.52; F1 score: 99.07). In addition to improved performance, the proposed methodology is simple and does not require more computational resources and time. the EnProCell is available at https://github.com/umar1196/EnProCell.

This paper presents a comprehensive analysis of the double-differential $p_T$ distributions of charged particles in twelve distinct pseudorapidity regions of equal width in $pp$ collisions at center-of-mass energies of 0.9, 2.36, and 7 TeV. Utilizing the modified Tsallis function with mean transverse flow velocity, our study demonstrates a very good agreement between experimental data and the model employed. The fit quality is consistently high across all $p_T$ ranges, as assessed by Data/Fit panels accompanying each plot. Extracted parameters, including kinetic freeze-out temperature ($T_0$), transverse flow velocity ($\beta_T$), non-extensivity parameter ($q$) and mean transverse momentum $\langle p_T \rangle$ dependencies are shown on pseudorapidity ($\eta$) and collision energy ($\sqrt{s}$). $T_0$, $\beta_T$ and $\langle p_T \rangle$ exhibit a decreasing trend with increasing $\eta$ due to lower in energy transfer along high $\eta$ regions, while they show a heightened sensitivity to $\sqrt{s}$. $q$ increases with $\eta$, indicating a closer thermal equilibrium in mid-$\eta$ particles. The paper also explores correlations among these parameters, emphasizing relationships between $T_0$, $\beta_T$, $q$ and $\langle p_T \rangle$. Our study provides valuable insights into the thermal and dynamic characteristics of high-energy proton-proton collisions, contributing to the broader understanding of the bulk properties of nuclear matter produced in these interactions., Comment: 7 pages, 3 figures

Significant progress has been made in advancing large multimodal conversational models (LMMs), capitalizing on vast repositories of image-text data available online. Despite this progress, these models often encounter substantial domain gaps, hindering their ability to engage in complex conversations across new domains. Recent efforts have aimed to mitigate this issue, albeit relying on domain-specific image-text data to curate instruction-tuning data. However, many domains, such as agriculture, lack such vision-language data. In this work, we propose an approach to construct instruction-tuning data that harnesses vision-only data for the agriculture domain. We utilize diverse agricultural datasets spanning multiple domains, curate class-specific information, and employ large language models (LLMs) to construct an expert-tuning set, resulting in a 70k expert-tuning dataset called AgroInstruct. Subsequently, we expert-tuned and created AgroGPT, an efficient LMM that can hold complex agriculture-related conversations and provide useful insights. We also develop AgroEvals for evaluation and compare {AgroGPT's} performance with large open and closed-source models. {AgroGPT} excels at identifying fine-grained agricultural concepts, can act as an agriculture expert, and provides helpful information for multimodal agriculture questions. The code, datasets, and models are available at https://github.com/awaisrauf/agroGPT.

Autonomous unmanned aerial vehicle (UAV) swarm networks (UAVSNs) can effectively execute surveillance, connectivity, and computing services to ground users (GUs). These missions require trajectory planning, UAV-GUs association, task offloading, next-hop selection, and resources such as transmit power, bandwidth, caching, and computing allocation to improve network performances. Owing to the highly dynamic topology, limited resources, and non-availability of global knowledge, optimizing network performance in UAVSNs is very intricate. Hence, it requires an adaptive joint optimization framework that can tackle both discrete and continuous decision variables to ensure optimal network performance under dynamic constraints. Multi-agent deep reinforcement learning-based adaptive actor-critic framework can efficiently address these problems. This paper investigates the recent evolutions of actor-critic frameworks to deal with joint optimization problems in UAVSNs. In addition, challenges and potential solutions are addressed as research directions.

Action detection and understanding provide the foundation for the generation and interaction of multimedia content. However, existing methods mainly focus on constructing complex relational inference networks, overlooking the judgment of detection effectiveness. Moreover, these methods frequently generate detection results with cognitive abnormalities. To solve the above problems, this study proposes a cognitive effectiveness network based on fuzzy inference (Cefdet), which introduces the concept of "cognition-based detection" to simulate human cognition. First, a fuzzy-driven cognitive effectiveness evaluation module (FCM) is established to introduce fuzzy inference into action detection. FCM is combined with human action features to simulate the cognition-based detection process, which clearly locates the position of frames with cognitive abnormalities. Then, a fuzzy cognitive update strategy (FCS) is proposed based on the FCM, which utilizes fuzzy logic to re-detect the cognition-based detection results and effectively update the results with cognitive abnormalities. Experimental results demonstrate that Cefdet exhibits superior performance against several mainstream algorithms on the public datasets, validating its effectiveness and superiority. Code is available at https://github.com/12sakura/Cefdet., Comment: The paper has been accepted by ACM MM. If you find this work helpful, please consider citing our paper. Zhe Luo, Weina Fu, Shuai Liu, Saeed Anwar, Muhammad Saqib, Sambit Bakshi, Khan Muhammad (2024) Cefdet: Cognitive Effectiveness Network Based on Fuzzy Inference for Action Detection, 32nd ACM International Conference on Multimedia, online first, 10.1145/3664647.3681226

We found the quasi inverse of qubit channels as a unitary map, $\mathcal{E}^i$, by minimizing the average trace distance between the input state to the channel and the output of the quasi inverse channel for arbitrary qubit channel $\mathcal{E}$ and for arbitrary input states. The channel $\mathcal{E}$ was assumed completely positive and trace-preserving. To find the quasi inverse for mixed states, we proposed an alternative definition of the quasi inverse based on the mean square of the trace distance (MSTD) of the mixed input state of the original channel and the output of the quasi inverse. The definition based on the trace distance allowed easy generalization of the quasi inverse to mixed input states. The quasi inverse of the Pauli, generalized amplitude damping, mixed unitary, and tetrahedron channels calculated based on trace distance agreed with the one computed using fidelity in the special case of input states being pure., Comment: 4 pages

Capitalizing on vast amount of image-text data, large-scale vision-language pre-training has demonstrated remarkable zero-shot capabilities and has been utilized in several applications. However, models trained on general everyday web-crawled data often exhibit sub-optimal performance for specialized domains, likely due to domain shift. Recent works have tackled this problem for some domains (e.g., healthcare) by constructing domain-specialized image-text data. However, constructing a dedicated large-scale image-text dataset for sustainable area of agriculture and livestock is still open to research. Further, this domain desires fine-grained feature learning due to the subtle nature of the downstream tasks (e.g, nutrient deficiency detection, livestock breed classification). To address this we present AgriCLIP, a vision-language foundational model dedicated to the domain of agriculture and livestock. First, we propose a large-scale dataset, named ALive, that leverages customized prompt generation strategy to overcome the scarcity of expert annotations. Our ALive dataset covers crops, livestock, and fishery, with around 600,000 image-text pairs. Second, we propose a training pipeline that integrates both contrastive and self-supervised learning to learn both global semantic and local fine-grained domain-specialized features. Experiments on diverse set of 20 downstream tasks demonstrate the effectiveness of AgriCLIP framework, achieving an absolute gain of 7.8\% in terms of average zero-shot classification accuracy, over the standard CLIP adaptation via domain-specialized ALive dataset. Our ALive dataset and code can be accessible at \href{https://github.com/umair1221/AgriCLIP/tree/main}{Github}.

We propose ClassroomKD, a novel multi-mentor knowledge distillation framework inspired by classroom environments to enhance knowledge transfer between student and multiple mentors. Unlike traditional methods that rely on fixed mentor-student relationships, our framework dynamically selects and adapts the teaching strategies of diverse mentors based on their effectiveness for each data sample. ClassroomKD comprises two main modules: the Knowledge Filtering (KF) Module and the Mentoring Module. The KF Module dynamically ranks mentors based on their performance for each input, activating only high-quality mentors to minimize error accumulation and prevent information loss. The Mentoring Module adjusts the distillation strategy by tuning each mentor's influence according to the performance gap between the student and mentors, effectively modulating the learning pace. Extensive experiments on image classification (CIFAR-100 and ImageNet) and 2D human pose estimation (COCO Keypoints and MPII Human Pose) demonstrate that ClassroomKD significantly outperforms existing knowledge distillation methods. Our results highlight that a dynamic and adaptive approach to mentor selection and guidance leads to more effective knowledge transfer, paving the way for enhanced model performance through distillation.

Random numbers are central to various applications such as secure communications, quantum key distribution theory (QKD), statistics, and other tasks. One of today's most popular generators is quantum random numbers (QRNGs). The inherent randomness and true unpredictability in quantum mechanics allowed us to construct QRNGs that are more accurate and useful than traditional random number generators. Based on different quantum mechanical principles, several QRNGs have already been designed. The primary focus of this paper is the generation and analysis of quantum random numbers based on radioactive decay. In the experimental set, two beta-active radioactive sources, cobalt-60 (Co60) and Strontium-90 (Sr 90), and an ST-360 counter with a Geiger-Muller (GM) tube are used to record the counts. The recorded data was then self-tested by entropy and frequency measurement. Moreover, popular testing technique, the National Institute of Science and Technology (NIST) randomness testing is used, to ensure that the guaranteed randomness meets security standards. The research provides the impact of the nature of the radioactive source, the distance between the counter and sources, and the recording time of the counts on generating quantum random numbers of radioactive QRNGs.

This paper reformulates cross-dataset human pose estimation as a continual learning task, aiming to integrate new keypoints and pose variations into existing models without losing accuracy on previously learned datasets. We benchmark this formulation against established regularization-based methods for mitigating catastrophic forgetting, including EWC, LFL, and LwF. Moreover, we propose a novel regularization method called Importance-Weighted Distillation (IWD), which enhances conventional LwF by introducing a layer-wise distillation penalty and dynamic temperature adjustment based on layer importance for previously learned knowledge. This allows for a controlled adaptation to new tasks that respects the stability-plasticity balance critical in continual learning. Through extensive experiments across three datasets, we demonstrate that our approach outperforms existing regularization-based continual learning strategies. IWD shows an average improvement of 3.60\% over the state-of-the-art LwF method. The results highlight the potential of our method to serve as a robust framework for real-world applications where models must evolve with new data without forgetting past knowledge.

Website accessibility is essential for inclusiveness and regulatory compliance. Although third-party advertisements (ads) are a vital revenue source for free web services, they introduce significant accessibility challenges. Leasing a website\'s space to ad-serving technologies like DoubleClick results in developers losing control over ad content accessibility. Even on highly accessible websites, third-party ads can undermine adherence to Web Content Accessibility Guidelines (WCAG). We conduct the first large-scale investigation of 430K website elements, including nearly 100K ad elements, to understand the accessibility of ads on websites. We seek to understand the prevalence of inaccessible ads and their overall impact on the accessibility of websites. Our findings show that 67% of websites experience increased accessibility violations due to ads, with common violations including Focus Visible and On Input. Popular ad-serving technologies like Taboola, DoubleClick, and RevContent often serve ads that fail to comply with WCAG standards. Even when ads are WCAG compliant, 27% of them have alternative text in ad images that misrepresents information, potentially deceiving users. Manual inspection of a sample of these misleading ads revealed that user-identifiable data is collected on 94% of websites through interactions, such as hovering or pressing enter. Since users with disabilities often rely on tools like screen readers that require hover events to access website content, they have no choice but to compromise their privacy in order to navigate website ads. Based on our findings, we further dissect the root cause of these violations and provide design guidelines to both website developers and ad-serving technologies to achieve WCAG-compliant ad integration.

Reducing the atmospheric haze and enhancing image clarity is crucial for computer vision applications. The lack of real-life hazy ground truth images necessitates synthetic datasets, which often lack diverse haze types, impeding effective haze type classification and dehazing algorithm selection. This research introduces the HazeSpace2M dataset, a collection of over 2 million images designed to enhance dehazing through haze type classification. HazeSpace2M includes diverse scenes with 10 haze intensity levels, featuring Fog, Cloud, and Environmental Haze (EH). Using the dataset, we introduce a technique of haze type classification followed by specialized dehazers to clear hazy images. Unlike conventional methods, our approach classifies haze types before applying type-specific dehazing, improving clarity in real-life hazy images. Benchmarking with state-of-the-art (SOTA) models, ResNet50 and AlexNet achieve 92.75\% and 92.50\% accuracy, respectively, against existing synthetic datasets. However, these models achieve only 80% and 70% accuracy, respectively, against our Real Hazy Testset (RHT), highlighting the challenging nature of our HazeSpace2M dataset. Additional experiments show that haze type classification followed by specialized dehazing improves results by 2.41% in PSNR, 17.14% in SSIM, and 10.2\% in MSE over general dehazers. Moreover, when testing with SOTA dehazing models, we found that applying our proposed framework significantly improves their performance. These results underscore the significance of HazeSpace2M and our proposed framework in addressing atmospheric haze in multimedia processing. Complete code and dataset is available on \href{https://github.com/tanvirnwu/HazeSpace2M} {\textcolor{blue}{\textbf{GitHub}}}., Comment: Accepted by ACM Multimedia 2024

Batteryless IoT devices, powered by energy harvesting, face significant challenges in maintaining operational efficiency and reliability due to intermittent power availability. Traditional checkpointing mechanisms, while essential for preserving computational state, introduce considerable energy and time overheads. This paper introduces Approxify, an automated framework that significantly enhances the sustainability and performance of batteryless IoT networks by reducing energy consumption by approximately 40% through intelligent approximation techniques. \tool balances energy efficiency with computational accuracy, ensuring reliable operation without compromising essential functionalities. Our evaluation of applications, SUSAN and Link Quality Indicator (LQI), demonstrates significant reductions in checkpoint frequency and energy usage while maintaining acceptable error bounds.

This study investigates the use of Large Language Models (LLMs) for improved depression detection from users social media data. Through the use of fine-tuned GPT 3.5 Turbo 1106 and LLaMA2-7B models and a sizable dataset from earlier studies, we were able to identify depressed content in social media posts with a high accuracy of nearly 96.0 percent. The comparative analysis of the obtained results with the relevant studies in the literature shows that the proposed fine-tuned LLMs achieved enhanced performance compared to existing state of the-art systems. This demonstrates the robustness of LLM-based fine-tuned systems to be used as potential depression detection systems. The study describes the approach in depth, including the parameters used and the fine-tuning procedure, and it addresses the important implications of our results for the early diagnosis of depression on several social media platforms., Comment: 16 pages

Optimizing Deep Learning-based Simultaneous Localization and Mapping (DL-SLAM) algorithms is essential for efficient implementation on resource-constrained embedded platforms, enabling real-time on-board computation in autonomous mobile robots. This paper presents SPAQ-DL-SLAM, a framework that strategically applies Structured Pruning and Quantization (SPAQ) to the architecture of one of the state-ofthe-art DL-SLAM algorithms, DROID-SLAM, for resource and energy-efficiency. Specifically, we perform structured pruning with fine-tuning based on layer-wise sensitivity analysis followed by 8-bit post-training static quantization (PTQ) on the deep learning modules within DROID-SLAM. Our SPAQ-DROIDSLAM model, optimized version of DROID-SLAM model using our SPAQ-DL-SLAM framework with 20% structured pruning and 8-bit PTQ, achieves an 18.9% reduction in FLOPs and a 79.8% reduction in overall model size compared to the DROID-SLAM model. Our evaluations on the TUM-RGBD benchmark shows that SPAQ-DROID-SLAM model surpasses the DROID-SLAM model by an average of 10.5% on absolute trajectory error (ATE) metric. Additionally, our results on the ETH3D SLAM training benchmark demonstrate enhanced generalization capabilities of the SPAQ-DROID-SLAM model, seen by a higher Area Under the Curve (AUC) score and success in 2 additional data sequences compared to the DROIDSLAM model. Despite these improvements, the model exhibits performance variance on the distinct Vicon Room sequences from the EuRoC dataset, which are captured at high angular velocities. This varying performance at some distinct scenarios suggests that designing DL-SLAM algorithms taking operating environments and tasks in consideration can achieve optimal performance and resource efficiency for deployment in resource-constrained embedded platforms., Comment: To appear at the 18th International Conference on Control, Automation, Robotics and Vision (ICARCV), December 2024, Dubai, UAE

Primary healthcare is a crucial strategy for achieving universal health coverage. South Asian countries are working to improve their primary healthcare system through their country specific policies designed in line with WHO health system framework using the six thematic pillars: Health Financing, Health Service delivery, Human Resource for Health, Health Information Systems, Governance, Essential Medicines and Technology, and an addition area of Cross-Sectoral Linkages. Measuring the current accessibility of healthcare facilities and workforce availability is essential for improving healthcare standards and achieving universal health coverage in developing countries. Data-driven surveillance approaches are required that can provide rapid, reliable, and geographically scalable solutions to understand a) which communities and areas are most at risk of inequitable access and when, b) what barriers to health access exist, and c) how they can be overcome in ways tailored to the specific challenges faced by individual communities. We propose to harness current breakthroughs in Earth-observation (EO) technology, which provide the ability to generate accurate, up-to-date, publicly accessible, and reliable data, which is necessary for equitable access planning and resource allocation to ensure that vaccines, and other interventions reach everyone, particularly those in greatest need, during normal and crisis times. This requires collaboration among countries to identify evidence based solutions to shape health policy and interventions, and drive innovations and research in the region.

Stunting detection is a significant issue in Indonesian healthcare, causing lower cognitive function, lower productivity, a weakened immunity, delayed neuro-development, and degenerative diseases. In regions with a high prevalence of stunting and limited welfare resources, identifying children in need of treatment is critical. The diagnostic process often raises challenges, such as the lack of experience in medical workers, incompatible anthropometric equipment, and inefficient medical bureaucracy. To counteract the issues, the use of load cell sensor and ultrasonic sensor can provide suitable anthropometric equipment and streamline the medical bureaucracy for stunting detection. This paper also employs machine learning for stunting detection based on sensor readings. The experiment results show that the sensitivity of the load cell sensor and the ultrasonic sensor is 0.9919 and 0.9986, respectively. Also, the machine learning test results have three classification classes, which are normal, stunted, and stunting with an accuracy rate of 98\%., Comment: 14 pages, pre-print

The demand of the Internet of Things (IoT) has witnessed exponential growth. These progresses are made possible by the technological advancements in artificial intelligence, cloud computing, and edge computing. However, these advancements exhibit multiple challenges, including cyber threats, security and privacy concerns, and the risk of potential financial losses. For this reason, this study developed a computationally inexpensive one-dimensional convolutional neural network (1DCNN) algorithm for cyber-attack classification. The proposed study achieved an accuracy of 99.90% to classify nine cyber-attacks. Multiple other performance metrices have been evaluated to validate the efficacy of the proposed scheme. In addition, comparison has been done with existing state-of-the-art schemes. The findings of the proposed study can significantly contribute to the development of secure intrusion detection for IIoT systems., Comment: 4 pages, 5 figures, 1 table, 29th International Conference on Automation and Computing

Deploying Multi-Modal Large Language Models (MLLMs) in healthcare is hindered by their high computational demands and significant memory requirements, which are particularly challenging for resource-constrained devices like the Nvidia Jetson Xavier. This problem is particularly evident in remote medical settings where advanced diagnostics are needed but resources are limited. In this paper, we introduce an optimization method for the general-purpose MLLM, TinyLLaVA, which we have adapted and renamed TinyLLaVA-Med. This adaptation involves instruction-tuning and fine-tuning TinyLLaVA on a medical dataset by drawing inspiration from the LLaVA-Med training pipeline. Our approach successfully minimizes computational complexity and power consumption, with TinyLLaVA-Med operating at 18.9W and using 11.9GB of memory, while achieving accuracies of 64.54% on VQA-RAD and 70.70% on SLAKE for closed-ended questions. Therefore, TinyLLaVA-Med achieves deployment viability in hardware-constrained environments with low computational resources, maintaining essential functionalities and delivering accuracies close to state-of-the-art models.

This study analyzes the Ames Housing Dataset using CatBoost and LightGBM models to explore feature importance and causal relationships in housing price prediction. We examine the correlation between SHAP values and EconML predictions, achieving high accuracy in price forecasting. Our analysis reveals a moderate Spearman rank correlation of 0.48 between SHAP-based feature importance and causally significant features, highlighting the complexity of aligning predictive modeling with causal understanding in housing market analysis. Through extensive causal analysis, including heterogeneity exploration and policy tree interpretation, we provide insights into how specific features like porches impact housing prices across various scenarios. This work underscores the need for integrated approaches that combine predictive power with causal insights in real estate valuation, offering valuable guidance for stakeholders in the industry.

Nematic quantum fluids appear in strongly interacting systems and break the rotational symmetry of the crystallographic lattice. In metals, this is connected to a well-known instability of the Fermi liquid-the Pomeranchuk instability. Using scanning tunneling microscopy, we identified this instability in a highly unusual setting: on the surface of an elemental topological metal, arsenic. By directly visualizing the Fermi surface of the surface state via scanning tunneling spectroscopy and photoemission spectroscopy, we find that the Fermi surface gets deformed and becomes elliptical at the energies where the nematic state is present. Known instances of nematic instability typically need van-Hove singularities or multi-orbital physics as drivers. In contrast, the surface states of arsenic are essentially indistinguishable from well-confined isotropic Rashba bands near the Fermi level, rendering our finding the first realization of Pomeranchuk instability of the topological surface state.

Retrieval-Augmented Generation (RAG) systems using large language models (LLMs) often generate inaccurate responses due to the retrieval of irrelevant or loosely related information. Existing methods, which operate at the document level, fail to effectively filter out such content. We propose LLM-driven chunk filtering, ChunkRAG, a framework that enhances RAG systems by evaluating and filtering retrieved information at the chunk level. Our approach employs semantic chunking to divide documents into coherent sections and utilizes LLM-based relevance scoring to assess each chunk's alignment with the user's query. By filtering out less pertinent chunks before the generation phase, we significantly reduce hallucinations and improve factual accuracy. Experiments show that our method outperforms existing RAG models, achieving higher accuracy on tasks requiring precise information retrieval. This advancement enhances the reliability of RAG systems, making them particularly beneficial for applications like fact-checking and multi-hop reasoning.

The rise of social media has amplified the spread of fake news, now further complicated by large language models (LLMs) like ChatGPT, which ease the generation of highly convincing, error-free misinformation, making it increasingly challenging for the public to discern truth from falsehood. Traditional fake news detection methods relying on linguistic cues also becomes less effective. Moreover, current detectors primarily focus on binary classification and English texts, often overlooking the distinction between machine-generated true vs. fake news and the detection in low-resource languages. To this end, we updated detection schema to include machine-generated news with focus on the Urdu language. We further propose a hierarchical detection strategy to improve the accuracy and robustness. Experiments show its effectiveness across four datasets in various settings.

Nuclear structure properties and weak interaction rates of neutron rich even even iron (Fe) isotopes (A = 50 70) are investigated using the Interacting Boson Model1 (IBM1) and the proton neutron Quasiparticle Random Phase Approximation (pnQRPA) model. The IBM1 is used for the calculation of energy levels and the B(E2) values of neutron rich Fe isotopes. Later their geometry was predicted within the potential energy formalism of the IBM1 model. Weak interaction rates on neutron rich nuclei are needed for the modeling and simulation of presupernova evolution of massive stars. In the current study, we investigate the possible effect of nuclear deformation on stellar rates of even even Fe isotopes. The pnQRPA model is applied to calculate the weak interaction rates of selected Fe isotopes using three different values of deformation parameter. It is noted that, in general, bigger deformation values led to smaller total strength and larger centroid values of the resulting Gamow Teller distributions. This later translated to smaller computed weak interaction rates. The current finding warrants further investigation before it may be generalized.The reported stellar rates are up to 4 orders of magnitude smaller than previous calculations and may bear astrophysical significance., Comment: 27 pages, 6 figures, 7 Tables

Semantic communication (SC) offers promising advancements in data transmission efficiency and reliability by focusing on delivering true meaning rather than solely binary bits of messages. However, privacy concerns in SC might become outstanding. Eavesdroppers equipped with advanced semantic coding models and extensive knowledge could be capable of correctly decoding and reasoning sensitive semantics from just a few stolen bits. To this end, this article explores utilizing knowledge to enhance data privacy in SC networks. Specifically, we first identify the potential attacks in SC based on the analysis of knowledge. Then, we propose a knowledge-assisted privacy preserving SC framework, which consists of a data transmission layer for precisely encoding and decoding source messages, and a knowledge management layer responsible for injecting appropriate knowledge into the transmission pair. Moreover, we elaborate on the transceiver design in the proposed SC framework to explain how knowledge should be utilized properly. Finally, some challenges of the proposed SC framework are discussed to expedite the practical implementation.