Abstract Data-driven domain adaptation methods have attracted significant attention in fault diagnosis. However, industrial deployment remains challenging due to scarce labeled data and diverse operating conditions. To tackle these challenges, a fault dynamics-assisted open-set domain adaptation (FD-ODA) was proposed for fault diagnosis across digital and physical domains. A fault dynamics model is built by the Augmented Lagrange multi-body dynamics approach to simulate typical bearing faults, including raceway defects and cage pillar fractures, and generate dataset of labeled vibration signals. The simulated signals are validated against experimental data obtained from test bench measurements. And to improve diagnostic accuracy, an open-set domain adaptation model is designed to reduce distribution shifts between virtual and real spaces. Experiments show that FD-ODA achieves superior performance in cross-domain open-set bearing fault diagnosis.

Meta-learners for few-shot weakly-supervised optic disc and cup segmentation on fundus images.

Information-theoretic continuously indexed domain adaptation network with wavelet-scale-wise convolution for fault diagnosis under continuously varying working conditions.

Domain adaptation methods have been extensively studied for rolling bearing fault diagnosis under various conditions. However, some existing methods only consider the one-way embedding of original space into a low-dimensional subspace without backward validation, which leads to inaccurate embeddings of data and poor diagnostic performance. In this paper, a rolling bearing fault diagnosis method based on multi-source domain joint structure preservation transfer with autoencoder (MJSPTA) is proposed. Firstly, similar source domains are screened by inter-domain metrics; then, the high-dimensional data of both the source and target domains are projected into a shared subspace with different projection matrices, respectively, during the encoding stage. Finally, the decoding stage reconstructs the low-dimensional data back to the original high-dimensional space to minimize the reconstruction accuracy. In the shared subspace, the difference between source and target domains is reduced through distribution matching and sample weighting. Meanwhile, graph embedding theory is introduced to maximally preserve the local manifold structure of the samples during domain adaptation. Next, label propagation is used to obtain the predicted labels, and a voting mechanism ultimately determines the fault type. The effectiveness and robustness of the method are verified through a series of diagnostic tests.

Aneurysmal subarachnoid hemorrhage (SAH) is a life-threatening neurological emergency with mortality rates exceeding 30%. While deep learning techniques show promise for automated SAH segmentation, their clinical application is limited by the scarcity of labeled data and challenges in cross-institutional generalization. Transfer learning from related hematoma types represents a potentially valuable but underexplored approach. Although Unet architectures remain the gold standard for medical image segmentation due to their effectiveness on limited datasets, Low-Rank Adaptation (LoRA) methods for parameter-efficient transfer learning have been rarely applied to convolutional neural networks in medical imaging contexts. The importance of SAH diagnosis and the time-intensive nature of manual annotation would benefit from automated solutions that can leverage existing multi-institutional datasets from more common conditions. We implemented a Unet architecture pre-trained on computed tomography scans from 124 traumatic brain injury patients across multiple institutions, then fine-tuned on 30 aneurysmal SAH patients from the University of Michigan Health System using 3-fold cross-validation. We developed a novel CP-LoRA method based on tensor canonical polyadic (CP) decomposition and introduced DoRA variants (DoRA-C, convDoRA, CP-DoRA) that decompose weight matrices into magnitude and directional components. We compared these approaches against existing LoRA methods (LoRA-C, convLoRA) and standard fine-tuning strategies across different modules on a multi-view Unet model. Performance was evaluated using Dice scores stratified by hemorrhage volume, with additional assessment of predicted versus annotated blood volumes. Transfer learning from traumatic brain injury to aneurysmal SAH demonstrated feasibility with all fine-tuning approaches achieving superior performance compared to no fine-tuning (mean Dice 0.410 ± 0.26). The best-performing traditional approach was decoding module fine-tuning (Dice 0.527 ± 0.20). LoRA-based methods consistently outperformed standard Unet fine-tuning, with DoRA-C at rank 64 achieving the highest overall performance (Dice 0.572 ± 0.17). Performance varied by hemorrhage volume, with all methods showing improved accuracy for larger volumes (Dice 0.682-0.694 for volumes > 100 mL vs. Dice 0.107-0.361 for volumes < 25 mL). CP-LoRA achieved comparable performance to existing methods while using significantly fewer parameters. Over-parameterization with higher ranks (64-96) consistently yielded better performance than strictly low-rank adaptations. This study demonstrates that transfer learning between hematoma types is feasible and that LoRA-based methods significantly outperform conventional Unet fine-tuning for aneurysmal SAH segmentation. The novel CP-LoRA method offers parameter efficiency advantages, while DoRA variants provide superior segmentation accuracy, particularly for small-volume hemorrhages. The finding that over-parameterization improves performance challenges traditional low-rank assumptions and suggests clinical applications may benefit from higher-rank adaptations. These results support the potential for automated SAH segmentation systems that leverage large multi-institutional traumatic brain injury datasets, potentially improving diagnostic speed and consistency when specialist expertise is unavailable.

One of the most concerning environmental issues that has attracted global attention is climate change and the problem of global warming. Climate change, along with its severe and detrimental impacts on the environment, threatens the fundamental elements of people's lives worldwide. The aim of this study is to analyze methods for enhancing farmers' resilience to cope with and adapt to the negative effects of climate change, and to provide recommendations for the localization of resilience-building strategies in the rural areas of Markazi Province. To achieve this objective, preliminary studies were conducted using documentary research methods. Subsequently, through careful observation of local lifestyles, semi-structured in-depth interviews lasting 2 to 3 hours, focused group discussions (FGD), and face-to-face conversations, the relationship between the localization of strategies and the enhancement of resilience among farmers was analyzed. The results of the analyses indicated that farmers’ use of localized methods positively influenced resilience in environmental, economic, social, and cultural dimensions, and could lead to greater adaptation to climate change and increased resilience. Consequently, it is recommended that resilience training in environmental, economic, social, and cultural aspects be provided by social workers, who are considered one of the main pillars of social support in communities. Instead of offering uniform and general training, the focus should be placed on the specific needs, conditions, and experiences of each region and farmer, thereby benefiting from the advantages of localization. This research can serve as a scientific basis for implementing the localization of methods in climate change adaptation policies, with the aim of enhancing the effectiveness of resilience training among farmers in the rural areas of Markazi Province.

Collaborative robotics in agriculture is designed to automate labor-intensive processes. In contrast to traditional autonomous systems, collaborative multi-agent robotic systems require active interaction between robots and human operators. This interaction creates the need for new methods for coordination, adaptation, and safety assurance in uncertain and dynamically changing environments. ( Research purpose ) The study aims to develop both theoretical and practical approaches to modeling the behavior and control of collaborative multi-agent robotic systems. The primary objective is to ensure efficient task allocation, coordinated agent behavior, and safe human-robot interaction during fruit harvesting operations. ( Materials and methods ) To achieve these objectives, the study employed methods from game theory, machine learning, and risk-aware control. A mathematical model was developed to describe the interactions among agents, incorporating the probabilistic nature of the environment and the involvement of a human operator. The proposed solutions were validated through a combination of numerical simulations and experimental data collected from a testbed replicating real-world agricultural scenarios. ( Results and discussion ) Algorithms were developed to enable coordination, adaptation, and dynamic task redistribution within the collaborative multi-agent robotic system. These algorithms demonstrated robustness against sensor inaccuracies, communication delays, and external disturbances typical of agricultural settings. Special attention was given to the system’s ability to adapt to human operator inputs, including task prioritization and context-sensitive interaction strategies. Simulation results showed enhanced system performance, characterized by more balanced task distribution among robots, reduced conflict during joint operations, and minimized idle time. Safety metrics also improved, including a reduction in collision risks and fewer incorrect responses to the presence of human operators in the work area. ( Conclusions ) The developed models and algorithms provide a foundation for the design of intelligent collaborative multi-agent robotic systems capable of adaptive and safe interaction in agricultural production. Their application can enhance the efficiency of automated harvesting processes while reducing reliance on manual labor.

This article offers a review on major linguistic works devoted to modern Khmer languages that were published in Russian, English, French and Khmer languages. A detailed description reflects the most thoroughly studied issues on phonetics, lexicology, word formation, morphology, and syntax of the Khmer language. It is noted that vocalism, the consonant system, and the structure of the Khmer syllable have been precisely investigated in phonetics. However, prosody has not been observed sufficiently yet. Lexicological examination focuses on the characterization of vocabulary from the standpoint of etymology, classification of foreign vocabulary by donor language, themes and method of adaptation, and systemic relationships of lexical units. Significant gaps are observed in the studies of onomastics, phraseology, and stylistics. Particular attention is paid to word formation methods: infixes and their meanings, prefixes, and semi-affixes. At the same time, some morphological research findings appear contradictory: parts of speech classifications vary among representatives of different linguistic schools. Syntax is thought to be the least depicted aspect of the Khmer language system. The analysis of functional, stylistic, and emotional-expressive properties of various language units, as well as the rules of their combinability when constructing phrases and sentences in the text is stated as important.

This study examines the cultural discount problem faced by entertainment contents such as films and television in cross-cultural communication under the background of globalization. Taking the Kung Fu Panda series as an example, it explores the strategies and effects of successfully breaking through cultural barriers. This research is of great significance for enriching the theory of cross-cultural communication and providing a practical framework for the cross-border dissemination of cultural products. This research mainly elaborates specifically from three aspects: cultural symbol selection and dissemination strategies, localization adaptation methods, and cross-cultural acceptance effects. The literature analysis method was applied to search for and read relevant materials and literature. The advantage of this method lies in its ability to analyze existing theoretical achievements and case materials well, which is conducive to systematically sorting out the mechanism of cultural dissemination and localization. The ultimate research objective of this study is to construct a cultural adaptation model suitable for the dissemination of cross-border film and television content. To achieve this goal, a multi-dimensional analysis was conducted on the use of symbols, narrative structure, language strategies, and audience responses in Kung Fu Panda.

Zigong Dinosaur Lantern Festival is an important representative of China's intangible cultural heritage, and its cultural connotation, communication paths, and cross-cultural adaptation methods deserve extensive attention from the academic community. This paper sorts out the cognitive framework of existing research on the cultural value of Zigong Lantern Festival, analyzes the stages and mechanisms of its international communication, and discusses three major challenges it faces in combination with cross-cultural communication theories: symbolic understanding deviation, unbalanced discourse structure, and insufficient ecological guarantee. The value system of Zigong Lantern Festival is built on the inheritance of historical folk customs, the practice of artistic innovation, and the empowerment of industrial ecology. Its international communication has experienced a development process from cultural exchange attempts, brand output expansion to digital innovation breakthroughs. In terms of cross-cultural adaptation, it is necessary to improve the actual communication effect by optimizing symbolic translation, constructing communication networks, and improving institutional guarantees.

ABSTRACT Cross-domain scene classification method is proposed to address the high cost of labeling remote sensing images and the poor generalization of supervised models on cross-domain data. Most existing convolutional neural network (CNN)-based domain adaptation methods rely on final-layer features or directly fuse multi-layer features for domain alignment, which may fail to capture multi-scale object information and introduce redundant features. To overcome these issues, a multi-level attention-guided contextual feature adaptive fusion network (MACFA-NET) is proposed. It incorporates a contextual feature injection module (CFIM) to enhance shallow features with deep semantic information while reducing redundancy. A global and local perception-guided attention module (GLPAM) is designed to capture complementary information flows between adjacent layers, promoting effective learning of domain-invariant representations. Additionally, a confidence-aware pseudo-label assignment strategy improves the reliability of pseudo-labels in the target domain. Experimental results demonstrate that MACFA-NET achieves superior cross-domain performance compared to existing methods.

Segmentation of infrared images is frequently essential in remote sensing image processing. However, infrared imagery poses significant challenges compared to conventional visible image segmentation, low resolution, insufficient data and poor labeling quality. The unsupervised domain adaptation method becomes an effective solution to this problem, which aims to enable a model trained on a source domain dataset with labeled information to adapt to another unlabeled target dataset. In this paper, the research of unsupervised domain adaptation method for semantic segmentation of remote sensing images is carried out based on deep learning. The training process of deep convolutional network requires a large amount of labeled data. However, remote sensing images are affected by differences in data acquisition location, time, light angle and other optical influences. To address this challenge, this paper proposes a remote sensing image domain adaptation semantic segmentation network that combines surface features and deep features. Photometric alignment is employed to intuitively reduce the domain gap and align surface features, while prototype-based classification and contrastive learning are used to align deep features. The experimental results demonstrate that this method achieves effective and reliable segmentation results on remote sensing image data from different scenes and improves the generalization ability of the semantic segmentation model.

Generally, a neural network is globally trained using the dataset provided in the training phase to optimize its parameters. The trained neural network is then used to make predictions for query instances in the inference phase. This global learning approach leads to a neural network that performs well universally across various query instances. However, it may overlook local structures in some low-density data regions, potentially degrading generalization performance in these regions. Although several test-time adaptation methods have been explored in recent years, they are typically designed for vision domains and are not intended for or do not readily transfer to tabular data. In this study, we propose a test-time local training method, specifically tailored for tabular data, to make the neural network better reflect the local structure around the query instance during the inference phase. Given a query instance, the proposed method finds the nearest neighbors of that instance from the training dataset. It then localizes the globally trained neural network by fine-tuning with these nearest neighbors to better accommodate the local structure around the query instance. The localized neural network is finally used to make a prediction for the query instance. Through experiments conducted on tabular benchmark datasets for regression and classification tasks, we demonstrate that the proposed method significantly enhances the generalization ability of neural networks.

This paper examines the emergence and evolution of simil sync in the Italian audiovisual sector, which originated in the mid-2000s as a spontaneous adaptation of voiceover techniques for non-scripted genres on thematic channels and OTT platforms. It traces changes in terminology, technical development, and recognition within the industry, also influenced by National Collective Labor Agreements (CCNLs). The study surveys Italian TV genres that shaped adaptation styles and examines the shifting role of non-automated adaptors and dubbing actors amid growing AI integration. Given the status of simil sync as a low-quality adaptation and niche modality, the concluding section explores potential future scenarios, with particular attention to the impact of AI-driven tools. Three video editing platforms – Dubverse, Vidnoz, and Wondershare Filmora 14 – were tested on a short clip containing the idiom “hit the books,” evaluated on (1) qualitative and (2) quantitative sync, (3) audio-visual/multimodal consistency, (4) translation accuracy, and (5) voice naturalness – prioritizing an accessible and audience-friendly viewing experience. Lay summary For decades, dubbing has been the predominant audiovisual translation (AVT) modality in Italy. Over time, however, alternative adaptation methods have emerged, with varying degrees of audience acceptance. This study focuses on “linear sync”, a term introduced in late 2023 to replace Sincronismo Ritmico Non Labiale (SRNL), referring to a half-synchronized adaptation process also known as “semi-sinc” (Sileo, 2018a, 2020) or “simil sync”. Officially recognized in the 2017 industry agreement, “linear sync” emerged in the early 2000s as a more efficient and cost-effective solution for adapting factual programs (Sileo, 2020). While simil sync, like standard dubbing, aims to match the duration of utterances in the target language with the original, it does not prioritize lip synchronization. This distinction has sparked debate within the AVT community: many professionals argue that it is less challenging than traditional dubbing, whereas linear sync adaptors present a different perspective on the process. Furthermore, within the Italian dubbing industry and among audiences, the AVT mode holds a complex reputation. Often regarded as a “stepchild” of standard dubbing (Barra et al., 2020), this hybrid transfer mode is sometimes criticized for its perceived simplicity and association with non-fiction content (Antoniazzi &amp; Barra, 2020; Sileo, 2020). This paper opens with an overview of linear sync or simil sync, tracing its development from inception to its current status in Italian AVT. Building on this historical context and previous research, we conclude with insights into the future path of simil sync in terms of its potential for development and appreciation in the Italian audiovisual landscape. This study wishes to contribute to our understanding of evolving AVT practices and the factors which may shape the future of audiovisual translation in Italy.

Abstract Climate change is one of the most prominent contemporary environmental phenomena that negatively impacts various sectors, especially the agricultural sector, which is the most affected, especially in arid and semi-arid regions such as Iraq. In this context, date palm cultivation is of great economic and social importance in Najaf Governorate. It is one of the most important strategic crops that farmers rely on as a primary source of income, in addition to its vital role in enhancing food security at the local and national levels. However, this cultivation faces increasing threats due to climate change, which has directly and indirectly affected palm productivity. This study aims to explore the relationship between climate changes, including temperature, rainfall, relative humidity, wind speed, and solar brightness. Using advanced statistical methods (using the ARDL model to analyze the relationship between climate variables and date production), climate and production data were collected over a period of twenty years. These data were analyzed using EViews software. The study covers date palm productivity in Iraq and the study area for the period from 2003 to 2023. The study will also analyze climate data for the period from 1993 to 2024. The study was based on an integrated scientific methodology that combined descriptive analysis with the use of questionnaires and interviews with decision-makers in the agricultural sector. It included distributing a questionnaire to a sample of 16 local decision-makers involved in the agricultural and environmental sectors. The study included directors of departments and divisions concerned with agriculture, water resources, environment, and climate. It also included secondary data collected from official sources, such as the Ministry of Agriculture (Agricultural Meteorology Center/Al-Mashkhab Station) and the Ministry of Transport (General Authority of Meteorology and Seismic Monitoring), in addition to international organizations specializing in climate and agriculture, such as the World Meteorological Organization (WMO), the Food and Agriculture Organization (FAO), and the International Conference on Climate Change (IPCC) The questionnaire included three main topics: the impact of climate change on agricultural production, adaptation strategies and associated challenges, and the role of government agencies and relevant institutions in supporting agriculture affected by these changes. Analysis results showed a significant decline in palm productivity over the past five years. Researchers indicated that the main reason behind this is the continuous rise in temperatures, especially during the flowering and fruiting months, which leads to fruit drop or incomplete ripening. Regarding adaptation, the study showed that most farmers lack the financial resources to purchase adaptation supplies, such as drip irrigation networks, shade covers, and sufficient fertilizers, It also revealed a clear lack of government plans and policies dedicated to adapting to climate change in palm cultivation. Based on these findings, the study recommends the urgent adoption of clear national and local plans to adapt to climate change. It also calls for strengthening farmers’ capacities through focused training and awareness programmes, and providing financial and technical support for adopting modern adaptation technologies. In addition, it is recommended to create integrated climate and agricultural databases that contribute to the formulation of future agricultural policies. The study also emphasizes the importance of incorporating climate change into agricultural strategic planning and strengthening partnerships between the public and private sectors and research institutions to address the negative impacts of climate change on date palm cultivation.

Improving Alzheimer's disease diagnosis by hyperspherical weighted adversarial learning in open set domain adaptation.

Continual source-free active domain adaptation for nasopharyngeal carcinoma tumor segmentation across multiple hospitals.

Abstract Unsupervised domain adaptation methods based on a single source domain assume that the target domain data does not contain label information, and diagnostic knowledge can be borrowed from a single source. These methods eliminate dependence on label information in target domain data, and as a result, they have been widely studied in the field of equipment fault diagnosis in recent years. However, in real-world scenarios, there are more challenging application cases: the diagnostic knowledge to be leveraged comes from multiple different operating conditions, or the target domain data lacks labeled information. Single-source domain methods struggle to effectively handle the differences between the source domain and the target domain. Additionally, existing diagnostic methods primarily focus on extracting common features between the source domain and target domain and aligning their distributions in a shared feature space, but they neglect the differences between source domains, between source domains and the target domain, and the decision boundaries between different fault categories. To address these issues, this paper proposes a domain adaptation fault diagnosis method based on dual-stage progressive alignment. First, several domain adaptation feature extractors are built to explicitly describe the structural variations between source domains, achieving feature adaptation in the first step, as opposed to standard multi-source transfer methods that handle all sources equally. In order to achieve feature adaptation in the second stage and more domain-adaptive results, a multi-source response classifier is designed to use the structural information of decision boundaries between source domains to guide the target domain based on feature-level alignment. Ultimately, the high-speed aviation bearing dataset is used for extensive experiments, and the outcomes show that the proposed approach excels at intelligent diagnostic tasks involving multiple source domains.

The deployment of unmanned aerial vehicles (UAVs) in cold climates poses substantial risks to aerodynamic performance and flight stability. Ice accumulation on the propellers elevates power consumption while diminishing the thrust necessary for sustaining flying agility. This study investigates the ice accretion process and its impact on the aerodynamic performance of a low-Reynolds-number UAV propeller during forward flight at speeds corresponding to maximum, optimal, and minimal thrust outputs, signifying differential effects on the accreted ice structures and the aerodynamic performance. Propellers functioning at higher thrust outputs (low advance ratios) during icing events produced lesser aerodynamic losses than those operating at elevated advance ratios with distinct ice structures, resulting in augmented boundary-layer losses, intensified vortex shedding, and increased flow turbulence in the wake region. Furthermore, vortex dynamics in the slipstream of iced propellers have been captured via a solution-based grid adaptation method, providing comprehensive insights into the vortex structures, a key parameter for evaluating the aerodynamic interactions in UAVs deployed for urban air mobility, revealing previously unexamined aspects of flow dynamics in iced UAVs.

A domain adaptation method to Defend Chinese textual adversarial attacks via prompt-tuning