Processing Pipeline Research Articles

A method to promote safe cycling powered by large language models and AI agents

This paper presents a novel information generation methodology to support safer cycling patterns in urban environments, leveraging for that Large Language Models (LLMs), AI-based agents, and open geospatial data. By processing multiple files containing previously computed urban risk levels and existing mobility infrastructure, which are generated by exploiting open data sources, our method exploits multi-layer data preprocessing procedures and prompt engineering to create easy-to-use, user-friendly assistive systems that are able to provide useful information concerning cycling safety. Through a well-defined processing pipeline based on Data Ingestion and Preparation, Agents Orchestration, and Decision Execution methodological steps, our method shows how to integrate open-source tools and datasets, ensuring reproducibility and accessibility for urban planners and cyclists. Moreover, an AI agent is also provided, which fully implements our method and acts as a proof-of-concept implementation. This paper demonstrates the effectiveness of our method in enhancing cycling safety and urban mobility planning.•A novel method that combines LLMs and AI agents is defined to enhance the processing of multi-domain open geospatial data, potentially promoting cycling safety.•It integrates urban risk data and cycling infrastructure for a more comprehensive understanding of cycling resources, which become accessible by textual or audio prompts.

Demystifying large language models in second language development research

Evaluating students' textual response is a common and critical task in language research and education practice. However, manual assessment can be tedious and may lack consistency, posing challenges for both scientific discovery and frontline teaching. Leveraging state-of-the-art large language models (LLMs), we aim to define and operationalize LLM-Surprisal, a numeric representation of the interplay between lexical diversity and syntactic complexity, and to empirically and theoretically demonstrate its relevance for automatic writing assessment and Chinese L2 (second language) learners’ English writing development. We developed an LLM-based natural language processing pipeline that can automatically compute text Surprisal scores. By comparing Surprisal metrics with the widely used classic indices in L2 studies, we extended the usage of computational metrics in Chinese learners’ L2 English writing. Our analyses suggested that LLM-Surprisals can distinguish L2 from L1 (first language) writing, index L2 development stages, and predict scores provided by human professionals. This indicated that the Surprisal dimension may manifest itself as critical aspects in L2 development. The relative advantages and disadvantages of these approaches were discussed in depth. We concluded that LLMs are promising tools that can enhance L2 research. Our showcase paves the way for more nuanced approaches to computationally assessing and understanding L2 development. Our pipelines and findings will inspire language teachers, learners, and researchers to operationalize LLMs in an innovative and accessible manner.

Representation of Social Determinants of Health terminology in medical subject headings: impact of added terms.

To enhance and evaluate the quality of PubMed search results for Social Determinants of Health (SDoH) through the addition of new SDoH terms to Medical Subject Headings (MeSH). High priority SDoH terms and definitions were collated from authoritative sources, curated based on publication frequencies, and refined by subject matter experts. Descriptive analyses were used to investigate how PubMed search details and best match results were affected by the addition of SDoH concepts to MeSH. Three information retrieval metrics (Precision, Recall, and F measure) were used to quantitatively assess the accuracy of PubMed search results. Pre- and post-update documents were clustered into topic areas using a Natural Language Processing pipeline, and SDoH relevancy assessed. Addition of 35 SDoH terms to MeSH resulted in more accurate algorithmic translations of search terms and more reliable best match results. The Precision, Recall, and F measures of post-update results were significantly higher than those of pre-update results. The percentage of retrieved publications belonging to SDoH clusters was significantly greater in the post- than pre-update searches. This evaluation confirms that inclusion of new SDoH terms in MeSH can lead to qualitative and quantitative enhancements in PubMed search retrievals. It demonstrates the methodology for and impact of suggesting new terms for MeSH indexing. It provides a foundation for future efforts across behavioral and social science research (BSSR) domains. Improving the representation of BSSR terminology in MeSH can improve PubMed search results, thereby enhancing the ability of investigators and clinicians to build and utilize a cumulative BSSR knowledge base.

SmartSPIM Pipeline: A Scalable Cloud-Based Image Processing Pipeline for Light-sheet Microscopy Data

SmartSPIM Pipeline: A Scalable Cloud-Based Image Processing Pipeline for Light-sheet Microscopy Data

ГИДРАВЛИЧЕСКИЙ РАСЧЕТ СИСТЕМЫ ПОДАЧИ ПЛАВЛЕНОГО СЫРА КУЛАЧКОВЫМ НАСОСОМ

The purpose of the study is to develop a calculation method for a continuous supply system for highly viscous food masses, taking into account the characteristics of the pipeline, cam pump and rheological parameters of the pumped food liquid. The paper considered a system for supplying processed cheese with rheological parameters at temperatures from 55 to 95 °C. As a result of hydraulic calculations using published data, it was found that 1) calculation of the effective dynamic viscosity μE using the Ostwald formula with the found parameter values yields results very close to the experimental data obtained from open sources; 2) the dependence of the expended power N on the pressure is linear, on the rotor speed n is parabolic, and on the dimensionless dynamic viscosity coefficient μ is power-law; 3) the dimensionless dynamic viscosity coefficient μ has different effects on the productivity Q and the power N: for large values of μ, the effect on Q decreases, while on N, on the contrary, it increases;4) with the growth of viscosity of the pumped liquid, the energy efficiency of the cam pump first improves (the hydraulic efficiency of the pump η increases, the specific energy consumption indicator E decreases), and then begins to deteriorate. The characteristic of the process pipeline (dependence of the required pressure on the flow rate of pumped processed cheese) with a given length L, local resistances, static pressure PC for three values of the internal diameter is constructed. A method is proposed that allows determining the operating parameters of the liquid food supply system at the operating point, taking into account the influence of its viscosity (using the example of pumping processed cheese).

IOP4, the Interactive Optical Photo-Polarimetric Python Pipeline

IOP4 is a pipeline to perform photometry and polarimetry analysis of optical data from Calar Alto (CAHA) and Sierra Nevada (OSN) observatories. IOP4 implements Object Relational Mapping to seamlessly integrate all information about the reduction and results in a database that can be used to query and plot results, flag data, and inspect the reduction process in an integrated fashion with the whole pipeline. It also ships with an already built-in web interface that can be used out of the box to browse the database and supervise all pipeline processes. It is built to ease debugging and inspection of data. Reduction from five different instruments are already implemented: RoperT90, AndorT90, DIPOL (at OSN 0.9 m telescope), AndorT150 (OSN 1.5 m telescope), and CAFOS (CAHA 2.2 m telescope). IOP4's modular design allows for easy integration of new observatories and instruments, and its results have already featured in several high-impact refereed publications. In this paper we describe the implementation and characteristics of IOP4.

Application of Event Cameras and Neuromorphic Computing to VSLAM: A Survey.

Simultaneous Localization and Mapping (SLAM) is a crucial function for most autonomous systems, allowing them to both navigate through and create maps of unfamiliar surroundings. Traditional Visual SLAM, also commonly known as VSLAM, relies on frame-based cameras and structured processing pipelines, which face challenges in dynamic or low-light environments. However, recent advancements in event camera technology and neuromorphic processing offer promising opportunities to overcome these limitations. Event cameras inspired by biological vision systems capture the scenes asynchronously, consuming minimal power but with higher temporal resolution. Neuromorphic processors, which are designed to mimic the parallel processing capabilities of the human brain, offer efficient computation for real-time data processing of event-based data streams. This paper provides a comprehensive overview of recent research efforts in integrating event cameras and neuromorphic processors into VSLAM systems. It discusses the principles behind event cameras and neuromorphic processors, highlighting their advantages over traditional sensing and processing methods. Furthermore, an in-depth survey was conducted on state-of-the-art approaches in event-based SLAM, including feature extraction, motion estimation, and map reconstruction techniques. Additionally, the integration of event cameras with neuromorphic processors, focusing on their synergistic benefits in terms of energy efficiency, robustness, and real-time performance, was explored. The paper also discusses the challenges and open research questions in this emerging field, such as sensor calibration, data fusion, and algorithmic development. Finally, the potential applications and future directions for event-based SLAM systems are outlined, ranging from robotics and autonomous vehicles to augmented reality.

GPU-accelerated OCT imaging: Real-time data processing and artifact suppression for enhanced monitoring of 3D bioprinted tissues and vascular-like networks

Optical coherence tomography (OCT) imaging technology has significant advantages in in situ and noninvasive monitoring of biological tissues. However, it still faces the following challenges: including data processing speed, image quality, and improvements in three-dimensional (3D) visualization effects. OCT technology, especially functional imaging techniques like optical coherence tomography angiography (OCTA), requires a long acquisition time and a large data size. Despite the substantial increase in the acquisition speed of swept source optical coherence tomography (SS-OCT), it still poses significant challenges for data processing. Additionally, during in situ acquisition, image artifacts resulting from interface reflections or strong reflections from biological tissues and culturing containers present obstacles to data visualization and further analysis. Firstly, a customized frequency domain filter with anti-banding suppression parameters was designed to suppress artifact noises. Then, this study proposed a graphics processing unit (GPU)-based real-time data processing pipeline for SS-OCT, achieving a measured line-process rate of 800[Formula: see text]kHz for 3D fast and high-quality data visualization. Furthermore, a GPU-based real-time data processing for CC-OCTA was integrated to acquire dynamic information. Moreover, a vascular-like network chip was prepared using extrusion-based 3D printing and sacrificial materials, with sacrificial material being printed at the desired vascular network locations and then removed to form the vascular-like network. OCTA imaging technology was used to monitor the progression of sacrificial material removal and vascular-like network formation. Therefore, GPU-based OCT enables real-time processing and visualization with artifact suppression, making it particularly suitable for in situ noninvasive longitudinal monitoring of 3D bioprinting tissue and vascular-like networks in microfluidic chips.

Maiolica seen by Vis–NIR hyperspectral imaging spectroscopy: the application of an ultraportable camera at the Museo Nazionale del Bargello

An ultraportable hyperspectral camera operating in the Vis–NIR range (400–1000 nm) was used in this study for the non-invasive analysis of a selection of Italian maiolica wares and sherds from the Museo Nazionale del Bargello in Florence, Italy. The studied objects included authentic archaeological sherds, nineteenth-century forgeries, and ceramics of uncertain origin. The primary aim of this study was to evaluate the efficacy of hyperspectral imaging (HSI) spectroscopy combined with multivariate analysis for examining highly refractive and three-dimensional glazed ceramic artefacts within the constraints of a museum environment, including imperfect lighting conditions. Two data processing pipelines were tested: one based on principal component analysis (PCA) with score plots and the other on spectral angle mapper (SAM) classification obtained through the ENVI Spectral Hourglass Wizard (ENVI-SHW). The analyses enabled the discrimination between original parts and/or restored parts and additions. Among others, a complex assemblage of original sherds, restored parts, and non-ceramic parts was observed in a Renaissance pastiche vase. Furthermore, the method enabled the identification of the various chromophores providing clues on the artistic colour palette. In particular, the identification of chromium in the blue and green glazes of two objects of uncertain chronology suggested that they were not made before the nineteenth century. The acquisition of a larger and more detailed database of HSI data on maiolica glazes is strongly encouraged to improve the applicability of this method in authenticity, art-historical and archaeological studies.

A user-friendly point cloud processing pipeline for interfacing PCL with YARP

A user-friendly point cloud processing pipeline for interfacing PCL with YARP

CLAY: A Controllable Large-scale Generative Model for Creating High-quality 3D Assets

In the realm of digital creativity, our potential to craft intricate 3D worlds from imagination is often hampered by the limitations of existing digital tools, which demand extensive expertise and efforts. To narrow this disparity, we introduce CLAY, a 3D geometry and material generator designed to effortlessly transform human imagination into intricate 3D digital structures. CLAY supports classic text or image inputs as well as 3D-aware controls from diverse primitives (multi-view images, voxels, bounding boxes, point clouds, implicit representations, etc). At its core is a large-scale generative model composed of a multi-resolution Variational Autoencoder (VAE) and a minimalistic latent Diffusion Transformer (DiT), to extract rich 3D priors directly from a diverse range of 3D geometries. Specifically, it adopts neural fields to represent continuous and complete surfaces and uses a geometry generative module with pure transformer blocks in latent space. We present a progressive training scheme to train CLAY on an ultra large 3D model dataset obtained through a carefully designed processing pipeline, resulting in a 3D native geometry generator with 1.5 billion parameters. For appearance generation, CLAY sets out to produce physically-based rendering (PBR) textures by employing a multi-view material diffusion model that can generate 2K resolution textures with diffuse, roughness, and metallic modalities. We demonstrate using CLAY for a range of controllable 3D asset creations, from sketchy conceptual designs to production ready assets with intricate details. Even first time users can easily use CLAY to bring their vivid 3D imaginations to life, unleashing unlimited creativity.

Semi-autonomous sensor fusion-based strategy for unmanned aerial atmospheric surveillance system in open field environments

<p>This paper explores the convergence of semi-autonomous systems and sensor fusion for monitoring hazardous atmospheric substances in open and semi-confined environments such as open-pit mines. As the heart of this system, a universal and flexible device leveraging wireless technologies to collect and analyze large volumes of data, designed by the authors, is proposed. The article outlines the key components of the platform, emphasizing its potential for increasing personnel safety levels as stipulated by the international public exposure guidelines. In modern mines, it becomes crucial to monitor elevated concentrations of nitrogen and carbon oxides, as well as other pollutants after blasting and the toxic gas emissions produced by the heavy transportation equipment such as mining trucks and excavators. In this paper, industrial-grade electrochemical sensors are compared with price-affordable but less precise microelectrochemical ones. The performed experiments for lowering computational requirements show promising results. The integration of an ultrasonic anemometer enhances the system’s capabilities, contributing to a comprehensive understanding of atmospheric conditions. In parallel, the research discusses the role of computer vision in autonomous control systems, with a focus on the architecture and processing pipeline of the state-of-the-art system-on-module Kria by AMD. Advocating the potential of adaptive computing for enhanced efficiency in dynamic environments, the paper underlines the importance of the integrated approach in developing a semi-autonomous atmospheric surveillance system and highlights the impact of adaptive computing in dynamic scenarios. The main part of the measuring equipment and methods has been experimented with in real conditions in open-pit mines in Bulgaria. This is an initial phase of ongoing research aimed at serving as a foundation for future improvements and the elaboration of a fully autonomous prototype for hazardous substances evaluation in the atmosphere of open-pit mines.<strong></strong></p>

Real-Time Precision in 3D Concrete Printing: Controlling Layer Morphology via Machine Vision and Learning Algorithms

3D concrete printing (3DCP) requires precise adjustments to parameters to ensure accurate and high-quality prints. However, despite technological advancements, manual intervention still plays a prominent role in this process, leading to errors and inconsistencies in the final printed part. To address this issue, machine learning vision models have been developed and utilized to analyze captured images and videos of the printing process, detecting defects and deviations. The data collected enable automatic adjustments to print settings, improving quality without the need for human intervention. This work first examines various techniques for real-time and offline corrections. It then introduces a specialized computer vision setup designed for real-time control in robotic 3DCP. Our main focus is on a specific aspect of machine learning (ML) within this system, called speed control, which regulates layer width by adjusting the robot motion speed or material flow rate. The proposed framework consists of three main elements: (1) a data acquisition and processing pipeline for extracting printing parameters and constructing a synthetic training dataset, (2) a real-time ML model for parameter optimization, and (3) a depth camera installed on a customized 3D-printed rotary mechanism for close-range monitoring of the printed layer.

Leveraging textual information for social media news categorization and sentiment analysis.

The rise of social media has changed how people view connections. Machine Learning (ML)-based sentiment analysis and news categorization help understand emotions and access news. However, most studies focus on complex models requiring heavy resources and slowing inference times, making deployment difficult in resource-limited environments. In this paper, we process both structured and unstructured data, determining the polarity of text using the TextBlob scheme to determine the sentiment of news headlines. We propose a Stochastic Gradient Descent (SGD)-based Ridge classifier (RC) for blending SGDR with an advanced string processing technique to effectively classify news articles. Additionally, we explore existing supervised and unsupervised ML algorithms to gauge the effectiveness of our SGDR classifier. The scalability and generalization capability of SGD and L2 regularization techniques in RCs to handle overfitting and balance bias and variance provide the proposed SGDR with better classification capability. Experimental results highlight that our string processing pipeline significantly boosts the performance of all ML models. Notably, our ensemble SGDR classifier surpasses all state-of-the-art ML algorithms, achieving an impressive 98.12% accuracy. McNemar's significance tests reveal that our SGDR classifier achieves a 1% significance level improvement over K-Nearest Neighbor, Decision Tree, and AdaBoost and a 5% significance level improvement over other algorithms. These findings underscore the superior proficiency of linear models in news categorization compared to tree-based and nonlinear counterparts. This study contributes valuable insights into the efficacy of the proposed methodology, elucidating its potential for news categorization and sentiment analysis.

A Set of FMRI Quality Control Tools in AFNI: Systematic, in-depth, and interactive QC with afni_proc.py and more.

Quality control (QC) assessment is a vital part of FMRI processing and analysis, and a typically underdiscussed aspect of reproducibility. This includes checking datasets at their very earliest stages (acquisition and conversion) through their processing steps (e.g., alignment and motion correction) to regression modeling (correct stimuli, no collinearity, valid fits, enough degrees of freedom, etc.) for each subject. There are a wide variety of features to verify throughout any single-subject processing pipeline, both quantitatively and qualitatively. We present several FMRI preprocessing QC features available in the AFNI toolbox, many of which are automatically generated by the pipeline-creation tool, afni_proc.py. These items include a modular HTML document that covers full single-subject processing from the raw data through statistical modeling, several review scripts in the results directory of processed data, and command line tools for identifying subjects with one or more quantitative properties across a group (such as triaging warnings, making exclusion criteria, or creating informational tables). The HTML itself contains several buttons that efficiently facilitate interactive investigations into the data, when deeper checks are needed beyond the systematic images. The pages are linkable, so that users can evaluate individual items across a group, for increased sensitivity to differences (e.g., in alignment or regression modeling images). Finally, the QC document contains rating buttons for each "QC block," as well as comment fields for each, to facilitate both saving and sharing the evaluations. This increases the specificity of QC, as well as its shareability, as these files can be shared with others and potentially uploaded into repositories, promoting transparency and open science. We describe the features and applications of these QC tools for FMRI.

Prediction of oil pipeline process operating parameters based on mechanism and data mining

Abstract Precisely forecasting the operational characteristics of oil pipelines is essential for developing rational design, production, and operation strategies, as well as reducing energy consumption and saving energy. Due to significant disparities in the computation outcomes of conventional mechanism models and the inadequate performance of machine learning models when handling limited sample data, their conclusions likewise lack tangible significance. In this study, a novel physics-guided neural network (PGNN) model, which integrates mechanisms with machine learning models, is introduced. The proposed model incorporates essential physical intermediate factors that impact the temperature and pressure of oil pipelines as artificial neurons within the loss function. Additionally, an adaptive moment estimate approach is employed to optimize the parameters of the model. Through a comparative analysis of various models' predictive capabilities on an oil pipeline, it was shown that PGNN has the highest level of accuracy in forecasting pipeline temperature and pressure. Furthermore, PGNN demonstrates the ability to generate satisfactory prediction outcomes even with a limited sample size. Simultaneously, the predictive outcomes of PGNN exhibit a stronger correlation with variables that have a direct impact on temperature and pressure.

Spurious correlations in surface-based functional brain imaging.

The study of functional MRI data is increasingly performed after mapping from volumetric voxels to surface vertices. Processing pipelines commonly used to achieve this mapping produce meshes with uneven vertex spacing, with closer neighbours in sulci compared to gyri. Consequently, correlations between the fMRI time series of neighbouring sulcal vertices are stronger than expected. However, the causes, extent, and impacts of this bias are not well understood or widely appreciated. We explain the origins of these biases, and using in-silico models of fMRI data, illustrate how they lead to spurious results. The bias leads to leakage of anatomical cortical folding information into fMRI time series. We show that many common analyses can be affected by this "gyral bias", including test-retest reliability, fingerprinting, functional parcellations, regional homogeneity, and brain-behaviour associations. Finally, we provide recommendations to avoid or remedy this spatial bias.

An Adaptively Weighted Averaging Method for Regional Time Series Extraction of fMRI-based Brain Decoding.

Brain decoding that classifies cognitive states using the functional fluctuations of the brain can provide insightful information for understanding the brain mechanisms of cognitive functions. Among the common procedures of decoding the brain cognitive states with functional magnetic resonance imaging (fMRI), extracting the time series of each brain region after brain parcellation traditionally averages across the voxels within a brain region. This neglects the spatial information among the voxels and the requirement of extracting information for the downstream tasks. In this study, we propose to use a fully connected neural network that is jointly trained with the brain decoder to perform an adaptively weighted average across the voxels within each brain region. We perform extensive evaluations by cognitive state decoding, manifold learning, and interpretability analysis on the Human Connectome Project (HCP) dataset. The performance comparison of the cognitive state decoding presents an accuracy increase of up to 5% and stable accuracy improvement under different time window sizes, resampling sizes, and training data sizes. The results of manifold learning show that our method presents a considerable separability among cognitive states and basically excludes subject-specific information. The interpretability analysis shows that our method can identify reasonable brain regions corresponding to each cognitive state. Our study would aid the improvement of the basic pipeline of fMRI processing.

Robust pixel-wise illuminant estimation algorithm for images with a low bit-depth

Conventional illuminant estimation methods were developed for scenes with a uniform illumination, while recently developed methods, such as pixel-wise methods, estimate the illuminants at the pixel level, making them applicable to a wider range of scenes. It was found that the same pixel-wise algorithm had very different performance when applied to images with different bit-depths, with up to a 30% decrease in accuracy for images having a lower bit-depth. Image signal processing (ISP) pipelines, however, prefer to deal with images with a lower bit-depth. In this paper, the analyses show that such a reduction was due to the loss of details and increase of noises, which were never identified in the past. We propose a method combining the L1 loss optimization and physical-constrained post-processing. The proposed method was found to result in around 40% higher estimation accuracy, in comparison to the state-of-the-art DNN-based methods.

Combining Federated Machine Learning and Qualitative Methods to Investigate Novel Pediatric Asthma Subtypes: Protocol for a Mixed Methods Study.

Pediatric asthma is a heterogeneous disease; however, current characterizations of its subtypes are limited. Machine learning (ML) methods are well-suited for identifying subtypes. In particular, deep neural networks can learn patient representations by leveraging longitudinal information captured in electronic health records (EHRs) while considering future outcomes. However, the traditional approach for subtype analysis requires large amounts of EHR data, which may contain protected health information causing potential concerns regarding patient privacy. Federated learning is the key technology to address privacy concerns while preserving the accuracy and performance of ML algorithms. Federated learning could enable multisite development and implementation of ML algorithms to facilitate the translation of artificial intelligence into clinical practice. The aim of this study is to develop a research protocol for implementation of federated ML across a large clinical research network to identify and discover pediatric asthma subtypes and their progression over time. This mixed methods study uses data and clinicians from the OneFlorida+ clinical research network, which is a large regional network covering linked and longitudinal patient-level real-world data (RWD) of over 20 million patients from Florida, Georgia, and Alabama in the United States. To characterize the subtypes, we will use OneFlorida+ data from 2011 to 2023 and develop a research-grade pediatric asthma computable phenotype and clinical natural language processing pipeline to identify pediatric patients with asthma aged 2-18 years. We will then apply federated learning to characterize pediatric asthma subtypes and their temporal progression. Using the Promoting Action on Research Implementation in Health Services framework, we will conduct focus groups with practicing pediatric asthma clinicians within the OneFlorida+ network to investigate the clinical utility of the subtypes. With a user-centered design, we will create prototypes to visualize the subtypes in the EHR to best assist with the clinical management of children with asthma. OneFlorida+ data from 2011 to 2023 have been collected for 411,628 patients aged 2-18 years along with 11,156,148 clinical notes. We expect to complete the computable phenotyping within the first year of the project, followed by subtyping during the second and third years, and then will perform the focus groups and establish the user-centered design in the fourth and fifth years of the project. Pediatric asthma subtypes incorporating RWD from diverse populations could improve patient outcomes by moving the field closer to precision pediatric asthma care. Our privacy-preserving federated learning methodology and qualitative implementation work will address several challenges of applying ML to large, multicenter RWD data. DERR1-10.2196/57981.