Features For Classification Research Articles

MSBWO: A Multi-Strategies Improved Beluga Whale Optimization Algorithm for Feature Selection.

Feature selection (FS) is a classic and challenging optimization task in most machine learning and data mining projects. Recently, researchers have attempted to develop more effective methods by using metaheuristic methods in FS. To increase population diversity and further improve the effectiveness of the beluga whale optimization (BWO) algorithm, in this paper, we propose a multi-strategies improved BWO (MSBWO), which incorporates improved circle mapping and dynamic opposition-based learning (ICMDOBL) population initialization as well as elite pool (EP), step-adaptive Lévy flight and spiral updating position (SLFSUP), and golden sine algorithm (Gold-SA) strategies. Among them, ICMDOBL contributes to increasing the diversity during the search process and reducing the risk of falling into local optima. The EP technique also enhances the algorithm's ability to escape from local optima. The SLFSUP, which is distinguished from the original BWO, aims to increase the rigor and accuracy of the development of local spaces. Gold-SA is introduced to improve the quality of the solutions. The hybrid performance of MSBWO was evaluated comprehensively on IEEE CEC2005 test functions, including a qualitative analysis and comparisons with other conventional methods as well as state-of-the-art (SOTA) metaheuristic approaches that were introduced in 2024. The results demonstrate that MSBWO is superior to other algorithms in terms of accuracy and maintains a better balance between exploration and exploitation. Moreover, according to the proposed continuous MSBWO, the binary MSBWO variant (BMSBWO) and other binary optimizers obtained by the mapping function were evaluated on ten UCI datasets with a random forest (RF) classifier. Consequently, BMSBWO has proven very competitive in terms of classification precision and feature reduction.

Automatic classification method for water entry sounds in multi-interference environments

Abstract In marine environments with strong pulse interference, distinguishing water entry sounds from pulse interference with high accuracy presents a challenge. Conventional algorithms that only utilize the singular values of intrinsic mode functions (IMFs) as the classification feature can have many misclassified signals in noisy environments. To identify the water entry sound more precisely, this study introduces a classifier based on selected singular value and correlation coefficient features. Specifically, the correlation coefficient between each IMF and the original signal are used to select crucial singular values and eliminate uninformative components. Furthermore, the correlation coefficient also combines with singular values as classification features for water entry sounds. The classification experiment results of 2460 groups of water entry sound and pulse interference indicate that the proposed classifier improves classification accuracy by approximately 11.5% compared to using singular values alone and by approximately 2.4% compared to classification accuracy without eliminating uninformative IMFs.

Towards hybrid approach based SVM and Radiomics features for COVID-19 classification and segmentation

In the battle against the COVID-19 pneumonia outbreak, which is brought on by the coronavirus strain SARS-Cov-2, radiological chest exams, such as chest X-rays, are crucial. In order to understand the unique radiographic characteristics of COVID-19, this research looks into classification models to distinguish chest X-ray images based on Radiomics features. This study is performed with datasets composed of 136 segmented chest X-rays, which were used to train and test the categorization algorithms. First and second-order statistical texture characteristics were extracted from the right (R), left (L), superior, middle, and bottom lung zones for each lung side using the Pyradiomics collection. Data was divided into training (80%) and test (20%) groups for feature selection. After assessing the respective feature significance and confirmation accuracy, the most pertinent Radiomics features were chosen. A model of lung segmentation based grey level pixels was used to evaluate support vector machines (SVM) as possible classifiers (AUC = 83.7%). Our research reveals a preference for the upper lung zone and a preponderance of Radiomics feature selection in the right lung. Our future research will concentrate on COVID-19 categorization and segmentation for more precise forecast using a hybrid method based on SVM and Radiogenomics features.

Feature Engineering for Agile Requirement Management Using Semantic Analysis

Efficient management and prioritization of software requirements are critical challenges in agile projects, where requirements constantly evolve due to changing user needs, business goals, and regulatory updates. This paper explores the role of semantic feature extraction in enabling adaptive management strategies. Using the PROMISE Expanded Dataset and the Coquina Dataset, we employed TF-IDF weighted Word2Vec for advanced tokenization and feature extraction. Latent Dirichlet Allocation (LDA) was used to analyze how preprocessing steps like stop word removal impact topic representation, revealing that removing stop words improved topic specificity and coherence. To address class imbalance, Synthetic Minority Over-sampling Technique (SMOTE) was applied, enhancing the model's ability to handle underrepresented classes effectively. Principal Component Analysis (PCA) reduced the dimensionality of TF-IDF weighted Word2Vec embeddings from 100 features to 30, while Analysis of Variance (ANOVA) identified the most significant features for classification. The results obtained identified three features to have p-values below 0.05 as statistically significant, p-value = 0.0000605, p-value = 0.00000000469, and p-value = 0.0024. These extracted features could be used as input to training machine learning models for predicting and managing software requirements adaptively during agile development. With the reduction of ambiguities and sentiments of the user at the requirement phase, the development phase could be undertaken seamlessly with ease.

Time-Series Feature Selection for Solar Flare Forecasting

Solar flares are significant occurrences in solar physics, impacting space weather and terrestrial technologies. Accurate classification of solar flares is essential for predicting space weather and minimizing potential disruptions to communication, navigation, and power systems. This study addresses the challenge of selecting the most relevant features from multivariate time-series data, specifically focusing on solar flares. We employ methods such as Mutual Information (MI), Minimum Redundancy Maximum Relevance (mRMR), and Euclidean Distance to identify key features for classification. Recognizing the performance variability of different feature selection techniques, we introduce an ensemble approach to compute feature weights. By combining outputs from multiple methods, our ensemble method provides a more comprehensive understanding of the importance of features. Our results show that the ensemble approach significantly improves classification performance, achieving values 0.15 higher in True Skill Statistic (TSS) values compared to individual feature selection methods. Additionally, our method offers valuable insights into the underlying physical processes of solar flares, leading to more effective space weather forecasting and enhanced mitigation strategies for communication, navigation, and power system disruptions.

Augmenting BERT to Model Remediation Processes in Finnish Countermedia

This paper showcases a supervised machine learning classifier to bridge the gap between qualitative and quantitative research in media studies, leveraging recent advancements in data-driven approaches. Current machine learning methods make it possible to gain insights from large datasets that would be impractical to analyze with more traditional methods. Supervised document classification presents a good platform for combining specific domain knowledge and close reading with broader quantitative analysis. The study focuses on a dataset of 37 185 articles from the Finnish countermedia publication MVlehti, from which a randomly sampled 997 articles were annotated into three categories based on frame analysis. Contextual sequence representations from the finBERT language model, topic distributions from a trained topic model, and a structural, HTML-aware featureset developed in prior work are employed as classification features. The hypothesis that BERT-based embeddings could be improved upon by augmenting them with additional information is supported by recent promising results in natural language benchmarks and tasks (Peinelt, Nguyen, and Liakata 2020; Glazkova 2021). In our study, combining contextual embeddings with topics resulted in only marginal performance increases, and this improvement was observed mostly in minority classes. Despite this, potential future developments to achieve better classification performance are outlined. Based on the experiments, automated frame analysis with neural classifiers is possible, but the accuracy is not yet sufficient for inferences of high certainty.

Histopathologic classification and immunohistochemical features of papillary renal neoplasm with potential therapeutic targets.

Papillary renal cell carcinoma (pRCC) is the second most common histological subtype of renal cell carcinoma and is considered a morphologically and molecularly heterogeneous tumor. Accurate classification and assessment of the immunohistochemical features of possible therapeutic targets are needed for precise patient care. We aimed to evaluate immunohistochemical features and possible therapeutic targets of papillary renal neoplasms. We collected 140 papillary renal neoplasms from three different hospitals and conducted immunohistochemical studies on tissue microarray slides. We performed succinate dehydrogenase B, fumarate hydratase, and transcription factor E3 immunohistochemical studies for differential diagnosis and re-classified five cases (3.6%) of papillary renal neoplasms. In addition, we conducted c-MET, p16, c-Myc, Ki-67, p53, and stimulator of interferon genes (STING) immunohistochemical studies to evaluate their pathogenesis and value for therapeutic targets. We found that c-MET expression was more common in pRCC (classic) (p = .021) among papillary renal neoplasms and Ki-67 proliferation index was higher in pRCC (not otherwise specified, NOS) compared to that of pRCC (classic) and papillary neoplasm with reverse polarity (marginal significance, p = .080). Small subsets of cases with p16 block positivity (4.5%) (pRCC [NOS] only) and c-Myc expression (7.1%) (pRCC [classic] only) were found. Also, there were some cases showing STING expression and those cases were associated with increased Ki-67 proliferation index (marginal significance, p = .063). Our findings suggested that there are subsets of pRCC with c-MET, p16, c-MYC, and STING expression and those cases could be potential candidates for targeted therapy.

Attribute disentanglement and re-entanglement for generalized zero-shot learning

The key challenge in zero-shot learning is inferring latent semantic knowledge between visual and attribute features of seen classes to achieve knowledge transfer to unseen classes. To address the limitation that local attribute features can only ensure attribute-level recognition rather than classification of an entire class, some methods incorporate global information into the process or results of local features extraction for classification. However, these approaches have not effectively addressed the issue. To address these issues, we propose an Attribute Disentanglement and Re-entanglement for Generalized Zero-Shot Learning. Our model no longer implicitly or explicitly incorporates global information into local attribute features for classification. Instead, we adjust local attribute features to make them more suitable for classification in re-entanglement phase, while ensuring the correct extraction of these features in disentanglement phase. We employ appropriate optimization loss functions and achieve significant improvements on three challenging benchmark datasets. Compared to other similar methods, our model exhibits strong competitiveness.

Imbrication fabric as a diagnostic feature for the genetic classification of gravels deposited by fluid‐gravity versus sediment‐gravity subaerial flows

AbstractGravel transport in subaerial environments occurs through different flows that are conveniently classified as debris flows, debris floods and water flows based on their distinct morpho‐sedimentary dynamics and different implications for geomorphic hazard. Because distinctive features allowing gravelly sedimentary bodies to be ascribed to related genetic process are still a matter of discussion, this study aims to establish whether imbrication fabric represents a sedimentological fingerprint potentially applicable towards a more robust genetic classification of gravels. We analysed the fabric of 1007 imbricated clasts from modern and ancient deposits. Our results highlight statistically significant differences between imbrication fabrics in gravels deposited by different flows. Particles imbricated by water flows are typified by low imbrication angles (median of 35°) and elongated clasts oriented perpendicular to the flow. In contrast, debris‐flow gravels exhibit high imbrication angles (median of 65°) and elongated clasts oriented parallel to the flow. Debris‐flood deposits display elongated clasts both parallel and transverse to the main flow and intermediate values of imbrication angle (median of 47°). We propose that imbrication angles result from the combination of stability‐driven selection—a process acting under tractional transport and promoting the remobilization of high‐angle imbrication fabrics—and shear‐stress‐driven overriding—a mechanism leading to the formation of the higher imbrication angles—with the first dominating in water flows and the latter being effective in mass transport processes. The progressive change in imbrication fabrics from fluid‐gravity to sediment‐gravity flow deposits offers easily quantifiable sedimentological evidence to help in distinguishing genetic processes that contribute to the accumulation of gravels in alluvial and colluvial settings. Analysis of imbrication fabric can add valuable information, particularly as regards the classification of (1) coarse deposits in stratigraphic records and (2) modern debris flood deposits.

Holoprosencephaly spectrum: an up-to-date overview of classification, genetics and neuroimaging.

Holoprosencephaly (HPE) is a complex forebrain congenital malformation with widely variable outcomes. It represents a disorder of ventral induction, which begins in the fifth gestational week. Its main feature is forebrain cleavage failure, which prevents the brain complete division into right and left hemispheres, the normal development of midline structures, and the deep brain structure. Based on the severity of prosencephalic cleavage failure, three classic forms (lobar, semilobar, and alobar) were described, and subsequently, interhemispheric variant (syntelencephaly) and septopreoptic variants were proposed. This review proposes a practical imaging approach to diagnosing HPE spectrum disorders, allowing an easier recognition and earlier diagnosis, which is essential for prenatal care and adequate parental counseling. In addition, we intend to simplify the understanding of HPE through a didactic discussion, schematic illustrations, and descriptions of each entity's current classification and critical neuroimaging features, as well as the main differential diagnosis of HPE.

Extraction of Garlic in the North China Plain Using Multi-Feature Combinations from Active and Passive Time Series Data

Garlic constitutes a significant small-scale agricultural commodity in China. A key factor influencing garlic prices is the planted area, which can be accurately and efficiently determined using remote sensing technology. However, the spectral characteristics of garlic and winter wheat are easily confused, and the widespread intercropping of these crops in the study area exacerbates this issue, leading to significant challenges in remote sensing image analysis. Additionally, remote sensing data are often affected by weather conditions, spatial resolution, and revisit frequency, which can result in delayed and inaccurate area extraction. In this study, historical data were utilized to restore Sentinel-2 remote sensing images, aimed at mitigating cloud and rain interference. Feature combinations were devised, incorporating two vegetation indices into a comprehensive time series, along with Sentinel-1 synthetic aperture radar (SAR) time series and other temporal datasets. Multiple classification combinations were employed to extract garlic within the study area, and the accuracy of the classification results was systematically analyzed. First, we used passive satellite imagery to extract winter crops (garlic, winter wheat, and others) with high accuracy. Second, we identified garlic by applying various combinations of time series features derived from both active and passive remote sensing data. Third, we evaluated the classification outcomes of various feature combinations to generate an optimal garlic cultivation distribution map for each region. Fourth, we developed a garlic fragmentation index to assess the impact of landscape fragmentation on garlic extraction accuracy. The findings reveal that: (1) Better results in garlic extraction can be achieved using active–passive time series remote sensing. The performance of the classification model can be further enhanced by incorporating short-wave infrared bands or spliced time series data into the classification features. (2) Examination of garlic cultivation fragmentation using the garlic fragmentation index aids in elucidating variations in accuracy across the study area’s six counties. (3) Comparative analysis with validation samples demonstrated superior garlic extraction outcomes from the six primary garlic-producing counties of the North China Plain in 2021, achieving an overall precision exceeding 90%. This study offers a practical exploration of target crop identification using multi-source remote sensing data in mixed cropping areas. The methodology presented here demonstrates the potential for efficient, cost-effective, and accurate garlic classification, which is crucial for improving garlic production management and optimizing agricultural practices. Moreover, this approach holds promise for broader applications, such as nationwide garlic mapping.

Integrating the incident angle of main vessel of breast nodule with BI-RADS can improve the efficacy of breast malignancy evaluation

The aim of this study was to investigate the measurement of the incident angle of the main blood vessel, and the benefits of its integral with ultrasound malignant features of breast nodules for the assessment of breast malignancy based on BI-RADS. The incident angles of main blood vessels of 185 breast nodules in 185 patients who underwent breast nodule surgical excision or biopsy were quantitatively measured using color Doppler ultrasound from October 2022 to October 2023 in a tertiary hospital, and related data were collected and analyzed. Based on histopathology as the gold standard, the breast nodules were classified into benign and malignant groups. The incident angle values of both groups were compared, Receiver Operating Characteristic (ROC) curves were plotted, and the optimal cutoff value for distinguishing between benign and malignant breast nodules was determined. The malignancy risk of the breast nodules was assessed using the incident angle of the breast main vessel, BI-RADS classification, and a combination of ultrasound malignant features with the incident angle. The areas under the ROC curves (AUC) for each method were calculated and compared. The average incident angle of the main vessel of the breast nodule for the benign and malignant breast nodule groups was (41.47 ± 14.27)° and (22.65 ± 11.09)°, respectively, with a statistically significant difference (t = 10.027, P < 0.001). There was a significant negative correlation between the breast nodule vessel incident angle and histopathological malignancy (r = – 0.593, P < 0.001). The ROC curve and Youden index suggested that the optimal cutoff value for distinguishing between benign and malignant breast nodules using the vascular incident angle was 26.9°, with a sensitivity of 76.34%, specificity of 84.78%, positive predictive value of 83.53%, negative predictive value of 78.00%, and AUC of 0.853. The diagnostic performance of BI-RADS classification based on ultrasound malignant features of the breast nodules alone in assessing the malignancy risk of breast nodules showed a sensitivity of 78.50%, specificity of 92.39%, positive predictive value of 91.25%, negative predictive value of 79.95%, and AUC of 0.905. The integral of ultrasound malignant features and vascular incident angle for BI-RADS based assessment for breast nodule malignancy risk demonstrated a sensitivity of 90.32%, specificity of 89.13%, positive predictive value of 89.36%, negative predictive value of 90.11%, and AUC of 0.940. The differences in negative predictive value and AUC between ultrasound malignant features BI-RADS classification alone and the combination of ultrasound malignant features BI-RADS classification with the incident angle of the main vessel of the breast nodule were all statistically significant (x2 = 3.243, P = 0.042; Z = 2.955, P = 0.003). Measuring the incident angle of the main blood vessel of breast nodules and combining this measurement with ultrasound malignant features for BI-RADS classification can enhance the effectiveness of malignancy risk assessment of breast nodules, increase the negative predictive value, and potentially reduce unnecessary biopsies.

FNIRSNET: A multi-view spatio-temporal convolutional neural network fusion for functional near-infrared spectroscopy-based auditory event classification

Multi-view learning is a rapidly evolving research area focused on developing diverse learning representations. In neural data analysis, this approach holds immense potential by capturing spatial, temporal, and frequency features. Despite its promise, multi-view application to functional near-infrared spectroscopy (fNIRS) has remained largely unexplored. This study addresses this gap by introducing fNIRSNET, a novel framework that generates and fuses multi-view spatio-temporal representations using convolutional neural networks. It investigates the combined informational strength of oxygenated (HbO2) and deoxygenated (HbR) hemoglobin signals, further extending these capabilities by integrating with electroencephalography (EEG) networks to achieve robust multimodal classification. Experiments involved classifying neural responses to auditory stimuli with nine healthy participants. fNIRS signals were decomposed into HbO2/HbR concentration changes, resulting in Parallel and Merged input types. We evaluated four input types across three data compositions: balanced, subject, and complete datasets. Our fNIRSNET's performance was compared with eight baseline classification models and merged it with four common EEG networks to assess the efficacy of combined features for multimodal classification. Compared to baselines, fNIRSNET using the Merged input type achieved the highest accuracy of 83.22%, 81.18%, and 91.58% for balanced, subject, and complete datasets, respectively. In the complete set, the approach effectively mitigated class imbalance issues, achieving sensitivity of 83.58% and specificity of 95.42%. Multimodal fusion of EEG networks and fNIRSNET outperformed single-modality performance with the highest accuracy of 87.15% on balanced data. Overall, this study introduces an innovative fusion approach for decoding fNIRS data and illustrates its integration with established EEG networks to enhance performance.

Performance Comparison of Supervised Machine Learning Methods in Classifying Celestial Objects

In recent times, astronomy has entered a new era with rapidly growing data sources and advanced observation techniques. The construction of powerful telescopes has enabled the collection of spectral data from millions of celestial objects. However, the increasing number and variety of data have made it challenging to categorize these celestial objects. This study employs machine learning methods to address the fundamental problem of classifying stars, galaxies, and quasars in astronomy. The dataset underwent detailed preprocessing to identify effective features for classification. KNIME Analytics Platform was used for data analysis and visualization, facilitating rapid and efficient data analysis through its drag-and-drop interface. Among the machine learning methods used in our study—Decision Trees, Random Forest, and Naive Bayes—the highest accuracy rate of 97.86% was achieved with the Random Forest model. Notably, despite its lower overall performance compared to other models, the Naive Bayes classifier exhibited superior performance in distinguishing the STAR class, which is one of the study's interesting findings. Future studies aim to enhance model accuracy by using larger and more diverse datasets and exploring different machine learning algorithms. Additionally, the impact of deep learning methods on classification performance will be investigated.

Annotation-Based Study on Hypothetical Proteins in Bacteria Using Classification Features

Background: Hypothetical proteins (HPs) are those proteins whose functions are unknown; therefore, the present study was carried out to predict similarity-based functionality of HPs in selected bacteria Streptomyces coelicolor A3(2) and Neisseria meningitidis. Methods: Annotation-based approaches using Pfam, orthology, String, Bi-directional Best Blast Hit, PSLpred, Subloc, Cello, homology modeling, and computational tools were used in evaluating the functionality of HPs. Results: Thirty-one domains in both bacterial species were retrieved based on the E-value score and compared with bacterial species already existing in databases. Statistical analysis was duly done to check which features performed well Conclusion: Out of 31 HPs found in Streptomyces coleicolor strain A3(2), 14 domains were found to be uncharacterized in their functionality, while 2 uncharacterized domains in the case of Neisseria meningitidis were assigned a function on similarity-based approaches. The annotation of HPs is a challenge in bacteria as these are based on the similarity of proteins in other species.

A sound event detection support system for smart home based on “two-to-one” teacher–student learning

Sound event detection (SED) is a core technology in smart home projects that rely on detected sound events to trigger specific actions. SED systems face two major challenges: high labeling costs and complex acoustic environments. To reduce labeling costs, some semi-supervised systems extract both global and local features for classification. However, these methods treat global and local features equally, not accounting for their varying importance when recognizing different types of sound events. Furthermore, to address complex acoustic environments, some studies use multitask learning frameworks to introduce SED-related tasks as auxiliaries to improve detection performance. However, these methods fail to align tasks within the framework, leading to conflicting outputs that may limit system performance. To address these issues, in this paper we propose a “two-to-one” teacher-student learning based semi-supervised SED system. This system employs a gating mechanism to selectively enhance global and local features, improving adaptability to different types of sound events, and incorporates a cross-task alignment module to interact SED with related tasks, reducing the risk of performance degradation caused by conflicting outputs. Experimental results on two datasets demonstrate that our system achieves the best performance in all metrics, with EB-F1 scores of 48.1 % and 64.7 %, representing improvements of 15.3 % and 10.6 % over the baseline ConformerSED system, respectively. Our work offers an effective SED solution for smart home projects by providing a semi-supervised SED system that performs well while reducing labeling costs.

Tracking paddy rice acreage, flooding impacts, and mitigations during El Niño flooding events using Sentinel-1/2 imagery and cloud computing

The frequent occurrence of El Niño events, in the context of climate change, brings heavy precipitation and extreme heat, severely disrupting agricultural production. Previous efforts have focused on monitoring crop planting areas and evaluating affected crops during disasters. Nevertheless, a comprehensive analysis, including crop planting area mapping, crop damage assessment, and mitigation effectiveness throughout the entire course of a disaster, has been seldom addressed. In this study, we built a comprehensive framework to rapidly investigate the areas of early rice, the extent of flooding impacts, and the post-flood mitigations of early rice during the El Niño flooding event in a typical rice production region – Jiangxi Province in 2023. Early rice planting areas were first mapped by integrating 15-day time series gap-filled Sentinel-1/2 datasets using the Google Earth Engine (GEE) platform, based on a random forest classifier built with the 55 optimized training features. Then the flood-affected early rice map was produced by integrating the early rice planting areas and the Sentinel-1 images-based flood map. Finally, the post-flood newly planted rice fields were identified using the random forest algorithm and classification features from the Sentinel-1/2 images composited during four phenology phases of newly planted rice. The results showed the early rice planting area map, the flooding map, and the newly planted early rice map have overall accuracies of over 90 %. The early rice planting areas reached 120 × 104 ha, and an area of 3.60 × 104 ha (3 %) was flooded due to the heavy rain, and 3.43 × 104 ha flooded areas were newly planted, eventually mitigating the flooding impacts on the production of early rice. This study showcases the potential of all the available Sentinel-1/2 data, cloud computing, and well-established mapping algorithms for tracking rice areas, flooding impacts, and mitigations (i.e., after-flooding replanting) during extreme climate events. The established framework is expected to serve as an early warning system for agricultural adaptation to extreme climate events.

Hierarchical classification for improving parcel-scale crop mapping using time-series Sentinel-1 data

Parcel-scale crop classification utilizing time-series satellite observations is of significant importance in precision agriculture. The prior knowledge that crop types can be organized in a hierarchical tree structure is beneficial for improving crop classification. Moreover, the crop hierarchy aligns with the coarse-to-fine cognitive process of geographic scenes. Based on the crop hierarchy, this study developed a general hierarchical classification framework for enhancing crop mapping using time-series Sentinel-1 data. Central to this method is a deep-learning-based hierarchical classification model that explores and makes use of crop hierarchical knowledge. First, preprocessed Sentinel-1 data were geometrically overlaid onto farmland parcel maps to derive parcel-scale time-series features. Second, we constructed a hierarchical crop type system for study areas based on the crop phenology of labeled crop-type samples. Third, we developed a deep-learning-based hierarchical classification model to identify crop types for each parcel, to generate final crop-type classification maps. The proposed approach was further discussed and verified through the implementation of parcel-scale time-series crop hierarchical classifications in a study area in France with farmland parcel maps and time-series Sentinel-1 data. The classification results, indicating significant improvements greater than 4.0% in overall accuracy and 5.0% in F1 score over comparative methods, demonstrated the effectiveness of the proposed method in learning multi-scale time-series features for hierarchical crop classification utilizing Sentinel-1 data sequences.

Fusing spatial and frequency features for compositional zero-shot image classification

Compositional Zero-Shot Learning (CZSL) is a particular Zero-Shot Learning (ZSL) task that aims to utilize known concepts (e.g., states and objects) to identify novel state-object compositions for Image Classification. Previous works have primarily focused on disentangling concept compositions or exploring the complex interactions between the states and objects while neglecting the critical fact that the inference of many states and compositions is related to different frequency components, which should be analyzed from the global perspective. Therefore, we propose a Spatial-frequency Feature Fusion Network (SFFNet) to introduce a new branch that utilizes a frequency-domain filtering encoder to enhance key frequency components and capture non-local interactions adaptively. Besides, we also find that the widely used backbone in conventional CZSL settings behaves superior in perceiving local features. Thus, we construct a fusion block to combine both strengths to capture the local and non-local information. In addition, the traditional one-hot ground-truth distribution in the training phase does not reflect the accurate relationships between compositions, so we propose a composition-relation based label distribution regularization to encourage the model to actively learn the inner relationships between compositions, and extend this method to construct unseen composition pseudo distribution to further enhance the model’s generalization ability to unseen compositions. Extensive experiments and detailed analysis are conducted on three popular datasets, and the results show that our method can achieve state-of-the-art performance, which reveals its superiority in identifying novel compositions. Code is available at https://github.com/lisuyi/SFFNet_czsl.

Selective State DOT Lane Width Standards and Guidelines to Reduce Speeds and Improve Safety

This research investigates the lane width standards and guidelines implemented by various State Departments of Transportation (DOTs) to reduce vehicle speeds and enhance road safety. Lane width reduction is often perceived as a strategy to mitigate speed and improve safety. Still, its effectiveness and implications vary across different contexts, including regions, urban/rural settings, or other geometric design features. Drawing from interviews with five State DOTs and a review of their road design manuals, this study aims to identify suggested lane widths depending on the contexts, design exception process when narrowing or widening lane widths, and introduce representative before/after studies. The findings indicate that State DOTs tend to have lower recommended lane widths in urban areas than in rural areas. Moreover, lane width standards among these states vary due to several factors, including the geographical location of roadways (urban or rural areas), design or posted speeds, traffic volume, road classification, and geometric road design features. Design exceptions are required if the existing or proposed design element is incompatible with both AASHTO and department governing criteria. In conclusion, the findings will provide valuable insights and recommendations for policymakers, transportation planners, and road engineers to inform optimal lane width and decision-making processes.