Context-aware Research Articles

Pembentukan Akhlakul Karimah Remaja Muslim Desa Purworejo dalam Konteks Kesadaran Beragama

Purworejo village has its own uniqueness regarding the daily lives of Muslim teenagers in it. The idealism of daily patterns with full ethics shows the success of learning patterns in them. Words and actions in everyday life that are tied to ethics are interesting to review.The work is in the form of field research, using qualitative methods, a phenomenological approach, along with data triangulation, trying to find patterns of effective learning in the village. With a focus on a) the process of forming al-karimah morals in the context of religious awareness, b) identification of moral awareness in a religious context, c) the role of various elements in successful learning.The results of the research are a) the process of forming morals in religious awareness is to use a pattern of forming awareness by involving various parties, b) the success of formation is by having an ideal pattern of behavior, both in ethics and enthusiasm for learning, c) there are three important elements that play a role in this process, namely teachers/figures, parents, and also the community

Context Awareness Assisted Integration System for Land Vehicles

Accurate context awareness of land vehicles can assist integrated navigation systems. Motion behavior recognition is one context awareness of vehicles, based on which constraint information helps reduce the impact of short-term blockage of navigation signals on radio-frequency-based positioning systems. To improve the reliability of behavior recognition, we proposed a machine learning-based vehicle motion behavior recognition and constraint method (MLMRC). The machine learning-based recognition process is directly driven by raw data from low-cost MEMS-IMU, while the traditional threshold-based method relies on previous experience. Four categories of constraint information—sensor error calibration, velocity constraint, angle constraint, and position constraint—were constructed from the recognition results. Both the simulated vehicle experiment and real vehicle experiment demonstrate the performance of the MLMRC method. When there is a short-term blockage, the MLMRC method can reduce the positioning error from 17.2% to 38.3% compared with the traditional method, which effectively improves positioning accuracy and provides support for autonomous vehicles in complex urban environments.

Impact of contextual factors on new product development process: evidence from a large company in Thailand

This research investigates the impacts of contextual factors, including organization-, project-, product-, and market-related factors, on New Product Development (NPD) process and explores how these factors shape the NPD process of the company within a specific context—a large-sized company in Thailand. By interviewing 11 managers from a large-sized company in Thailand, the findings reveal different impact of the contextual factors on the company’s NPD process. Based on these findings, we develop a practical NPD model to overcome company-specific challenges and, consequently, offer valuable insights to enhance the NPD process in similar contexts. Importantly, our findings suggest that innovative companies should strategically integrate the route-to-market for radical innovation into their NPD process because it enhances the customer perception of the innovation’s value before commercialization, leading to the success of NPD. Our empirical evidence fills the gaps in the literature regarding the practical configuration of NPD processes in developing countries as well as large-sized companies, acknowledging the variation in the NPD process that depends on contextual factors where innovation occurs. It also provides detailed insights beyond most existing studies that investigate correlation of factors with the NPD process and offers practical implications for management involved in the NPD process, emphasizing the need for contextual awareness in formulating effective NPD strategies.

GGMNet: Pavement-Crack Detection Based on Global Context Awareness and Multi-Scale Fusion

Accurate and comprehensive detection of pavement cracks is important for maintaining road quality and ensuring traffic safety. However, the complexity of road surfaces and the diversity of cracks make it difficult for existing methods to accomplish this challenging task. This paper proposes a novel network named the global graph multiscale network (GGMNet) for automated pixel-level detection of pavement cracks. The GGMNet network has several innovations compared with the mainstream road crack detection network: (1) a global contextual Res-block (GC-Resblock) is proposed to guide the network to emphasize the identities of cracks while suppressing background noises; (2) a graph pyramid pooling module (GPPM) is designed to aggregate the multi-scale features and capture the long-range dependencies of cracks; (3) a multi-scale features fusion module (MFF) is established to efficiently represent and deeply fuse multi-scale features. We carried out extensive experiments on three pavement crack datasets. These were DeepCrack dataset, with complex background noises; the CrackTree260 dataset, with various crack structures; and the Aerial Track Detection dataset, with a drone’s perspective. The experimental results demonstrate that GGMNet has excellent performance, high accuracy, and strong robustness. In conclusion, this paper provides support for accurate and timely road maintenance and has important reference values and enlightening implications for further linear feature extraction research.

Artificial intelligence technologies in information and library systems

The effectiveness of innovative scientific and educational activities largely relies on the technologies of the library industry, including modern information and library systems. Academic and research libraries introduce new technologies in an effort to improve the quality of their services. Artificial intelligence is one of the key technologies for digital transformation of the library industry. The study is aimed at identifying trends in the use of artificial intelligence in information and library service technologies. The author applied the general dialectical method, theoretical analysis and generalization of the content of special, scientific and technical literature, regulatory documents in the field of digitalization of librarianship and artificial intelligence. The results of the study have revealed the role of the integration of artificial intelligence technologies into information and library systems, which lies in the fact that they contribute to the formation of national intellectual libraries, computerization of the routine library services, and structuring innovative academic digital space for the users. Using the example of integration of context awareness technology into some information and library systems, the ability of a digital system to make “reasonable” decisions to extract relevant information for learning and self-education is demonstrated, which also allows to find educational materials on the topic being studied, and to build explicit educational route for a user.

Enhancing Chatbot Intelligence Through Narrative Memory Structures

Through the use of narrative memory structures—a cutting-edge technique intended to provide chatbots context awareness and conversational coherence—we sought to improve the intelligence of chatbots. We carefully designed our process with the goal of immediately incorporating narrative data into the chatbot's memory architecture. In order to find narrative patterns, this required evaluating and classifying previous exchanges. The narrative patterns were then saved and used to guide the chatbot's answers. The incorporation of narrative memory structures into the chatbot resulted in a significant improvement in conversational coherence and relevance, as per our study's key findings. Through the use of its narrative memory, which stores earlier encounters, the chatbot shown an impressive capacity to sustain the conversation's flow and provide more contextually relevant replies. In addition, the chatbot demonstrated improved understanding of human intentions, correctly identifying underlying motives and preferences. In addition, comments produced by integrating narrative memory structures were highly customized in addition to being accurate. By utilizing the extensive history of previous exchanges, the chatbot was able to customize replies to each user's unique situation, building a closer relationship and rapport with them. All things considered; our study highlights how narrative memory processes can revolutionize chatbot intelligence. We can improve chatbots' conversational capabilities and enable more meaningful human-computer interactions by giving them the capacity to extract narrative information from previous conversations. This opens the door to a new era of chatbot designs that put the needs of users, contextuality, and empathy first, changing the game for conversational AI in the process.

CHATBOT USING LARGE LANGUAGE MODEL

The concept of Natural Language Processing has seen a remarkable advancement in the recent years. This remarkable advancement was particularly with the development of Large Language Models (LLM). Large Language Models are used to develop a human like conversations. This LLM is a part of Natural Language Processing which focuses on enabling computers to understand, interpret, and generate human language. The existing system of chatbots does not generate human like responses. The proposed system of chatbots uses the power of Large Language Models to generate more human like responses, providing the conversation in a natural way. By genereating human like respones, it will be in a natural way for the user. To enhance user experience, the chatbot uses a dynamic learning mechanism, by which it continuously adapt to user preferences and evolving conversational patterns. This system uses feedbacks from the users to refine its responses everytime.Moreover, the chatbot is designed with a multi-turn conversational context awareness, allowing it to maintain coherence and relevance throughout extended dialogues.The effectiveness of the proposed chatbot is evaluated through user testing, comparing its performance against traditional rule-based chatbots and existing conversational agents. This report explains about the usage of Large Language Models in the design and implementation of conversational chatbots. The outcomes of this research contribute to the advancement of intelligent chatbot systems, demonstrating the potential of large language models to significantly enhance conversational AI applications.

It takes a village: Guarding against sexual abuse in Australian educational settings

Reporting indicators of child sexual abuse is critical to intervention and prevention. However, guardianship and its mechanisms in educational settings have been empirically neglected. We address this gap by analysing a unique administrative dataset from an Australian jurisdiction that captures alleged improper sexual conduct by school employees against adolescents aged 13–17 years inclusive. We analysed 638 cases reported to an external oversight body in accordance with the New South Wales “reportable conduct” scheme between 2015 and 2019. These cases detailed information about the sources of allegations, including how allegations arose and who brought them forward. Results established school personnel and families as the key reporters and supported young people's potential as capable guardians. Guardianship was highly context-dependent and differed according to event, victim, and perpetrator characteristics. Recommendations for strengthening guardianship in schools are presented, including enabling reporting pathways for students, building contextual awareness through policy dissemination and training, and heeding indirect sources of information like rumours.

TinyNS: Platform-Aware Neurosymbolic Auto Tiny Machine Learning.

Machine learning at the extreme edge has enabled a plethora of intelligent, time-critical, and remote applications. However, deploying interpretable artificial intelligence systems that can perform high-level symbolic reasoning and satisfy the underlying system rules and physics within the tight platform resource constraints is challenging. In this paper, we introduce TinyNS, the first platform-aware neurosymbolic architecture search framework for joint optimization of symbolic and neural operators. TinyNS provides recipes and parsers to automatically write microcontroller code for five types of neurosymbolic models, combining the context awareness and integrity of symbolic techniques with the robustness and performance of machine learning models. TinyNS uses a fast, gradient-free, black-box Bayesian optimizer over discontinuous, conditional, numeric, and categorical search spaces to find the best synergy of symbolic code and neural networks within the hardware resource budget. To guarantee deployability, TinyNS talks to the target hardware during the optimization process. We showcase the utility of TinyNS by deploying microcontroller-class neurosymbolic models through several case studies. In all use cases, TinyNS outperforms purely neural or purely symbolic approaches while guaranteeing execution on real hardware.

Diagnostic efficacy of [68Ga]Ga-DOTA-GPFAPI-04 in patients with solid tumors in a head-to-head comparison with [18F]F-FDG: results from a prospective clinical study.

To identify the biodistribution and diagnostic performance of a novel fibroblast activation protein (FAP) targeted positron emission tomography (PET) tracer, [68Ga]Ga-DOTA-GPFAPI-04, in patients with solid tumors in a head-to-head comparison with [18F]F-FDG. Twenty-six patients histologically proven with cancers of nasopharyngeal (n = 5), esophagus (n = 5), gastro-esophagus (n = 1), stomach (n = 7), liver (n = 3), and colorectum (n = 5) were recruited for [68Ga]Ga-DOTA-GPFAPI-04 and [18F]F-FDG PET/CT scans on consecutive days. The primary endpoint was the diagnostic efficacy, with the histological diagnosis and the follow-up results selected as the gold standard. The secondary endpoint was the background uptake pattern. Two experienced nuclear medicine physicians who were blinded to the gold standard results while having essential awareness of the clinical context reviewed the images and labeled lesions by consensus for subsequent software-assisted lesion segmentation. Additionally, background organs were automatically segmented, assisted by artificial intelligence. Volume, mean, and maximum standard uptake values (SUVmean and SUVmax) of all segmentations were recorded. P < 0.05 was deemed as statistically significant. Significant glandular uptake of [68Ga]Ga-DOTA-GPFAPI-04 was detected in the thyroid, pancreas, and submandibular glands, while moderate uptake was observed in the parotid glands. The myocardium and myometrium exhibited 2-3 times higher uptake of the radiotracer than that of the background levels in blood and liver. A total of 349 targeted lesions, consisting of 324 malignancies and 25 benign lesions, were segmented. [68Ga]Ga-DOTA-GPFAPI-04 is more sensitive than [18F]F-FDG, especially for abdominopelvic dissemination (1.000 vs. 0.475, P < 0.001). Interestingly, [18F]F-FDG demonstrated higher sensitivity for lung metastasis compared to [68Ga]Ga-DOTA-GPFAPI-04 (0.845 vs. 0.682, P = 0.003). The high glandular uptake made it difficult to delineate lesions in close proximity and masked two metastatic lesions in these organs. Despite prominent glandular uptake, [68Ga]Ga-DOTA-GPFAPI-04 demonstrates favorable diagnostic performance. It is a promising probe scaffold for further development of FAP-targeted tumor theranostic agents.

A Novel Multi-Scale Feature Map Fusion for Oil Spill Detection of SAR Remote Sensing

The efficient and timely identification of oil spill areas is crucial for ocean environmental protection. Synthetic aperture radar (SAR) is widely used in oil spill detection due to its all-weather monitoring capability. Meanwhile, existing deep learning-based oil spill detection methods mainly rely on the classical U-Net framework and have achieved impressive results. However, SAR images exhibit high noise, blurry boundaries, and irregular shapes of target areas, as well as speckles and shadows, which lead to the loss of performance in existing algorithms. In this paper, we propose a novel network architecture to achieve more precise segmentation of oil spill areas by reintroducing rich semantic contextual information before obtaining the final segmentation mask. Specifically, the proposed architecture can re-fuse feature maps from different levels at the decoder end. We design a multi-convolutional layer (MCL) module to extract basic feature information from SAR images, and a feature extraction module (FEM) module further extracts and fuses feature maps generated by the U-Net decoder at different levels. Through these operations, the network can learn rich global and local contextual information, enable sufficient interaction of feature information at different stages, enhance the model’s contextual awareness, and improve its ability to recognize complex textures and blurry boundaries, thereby enhancing the segmentation accuracy of SAR images. Compared to many U-Net based segmentation networks, our method shows promising results and achieves state-of-the-art performance on multiple evaluation metrics.

Context-Aware System for Information Flow Management in Factories of the Future

The trends of the 21st century are challenging the traditional production process due to the reduction in the life cycle of products and the demand for more complex products in greater quantities. Industry 4.0 (I4.0) was introduced in 2011 and it is recognized as the fourth industrial revolution, with the aim of improving manufacturing processes and increasing the competitiveness of industry. I4.0 uses technological concepts such as Cyber-Physical Systems, Internet of Things and Cloud Computing to create services, reduce costs and increase productivity. In addition, concepts such as Smart Factories are emerging, which use context awareness to assist people and optimize tasks based on data from the physical and virtual world. This article explores and applies the capabilities of context-aware applications in industry, with a focus on production lines. In specific, this paper proposes a context-aware application based on a microservices approach, intended for integration into a context-aware information system, with specific application in the area of manufacturing. The manuscript presents a detailed architecture for structuring the application, explaining components, functions and contributions. The discussion covers development technologies, integration and communication between the application and other services, as well as experimental findings, which demonstrate the applicability and advantages of the proposed solution.

RIGA: Rotation-Invariant and Globally-Aware Descriptors for Point Cloud Registration.

Successful point cloud registration relies on accurate correspondences established upon powerful descriptors. However, existing neural descriptors either leverage a rotation-variant backbone whose performance declines under large rotations, or encode local geometry that is less distinctive. To address this issue, we introduce RIGA to learn descriptors that are Rotation-Invariant by design and Globally-Aware. From the Point Pair Features (PPFs) of sparse local regions, rotation-invariant local geometry is encoded into geometric descriptors. Global awareness of 3D structures and geometric context is subsequently incorporated, both in a rotation-invariant fashion. More specifically, 3D structures of the whole frame are first represented by our global PPF signatures, from which structural descriptors are learned to help geometric descriptors sense the 3D world beyond local regions. Geometric context from the whole scene is then globally aggregated into descriptors. Finally, the description of sparse regions is interpolated to dense point descriptors, from which correspondences are extracted for registration. To validate our approach, we conduct extensive experiments on both object- and scene-level data. With large rotations, RIGA surpasses the state-of-the-art methods by a margin of 8 ° in terms of the Relative Rotation Error on ModelNet40 and improves the Feature Matching Recall by at least 5 percentage points on 3DLoMatch.

Adjusting Clinical Plans Based on Social Context.

Social risk data collection is expanding in community health centers (CHCs). We explored clinicians' practices of adjusting medical care based on their awareness of patients' social risk factors-that is, changes they make to care plans to mitigate the potential impacts of social risk factors on their patients' care and health outcomes-in a set of Texas CHCs. Convergent mixed methods. Surveys/interviews explored clinician perspectives on adjusting medical care based on patient social risk factors. Survey data were analyzed with descriptive statistics; interviews were analyzed using thematic analysis and inductive coding. Across 4 CHCs, we conducted 15 clinician interviews and collected 97 surveys. Interviews and surveys overall indicated support for adjustment activities. Two main themes emerged: 1) clinicians reported making frequent adjustments to patient care plans based on their awareness of patients' social contexts, while simultaneously expressing concerns about adjustment; and 2) awareness of patients' social risk factors, and clinician time, training, and experience all influenced clinician adjustments. Clinicians at participating CHCs described routinely adjusting patient care plans based on their patients' social contexts. These adjustments were being made without specific guidelines or training. Standardization of adjustments may facilitate the contextualization of patient care through shared decision making to improve outcomes.

Broadening the coating applications of sustainable materials by reinforcing epoxidized corn oil with single-walled carbon nanotubes

In the global context of environmental awareness, the present research proposes a sustainable alternative to the widely used petroleum-based epoxy coatings. Epoxidized corn oil (ECO) was tested as potential matrix for advanced nanocomposite coating materials reinforced with 0.25 to 1 wt.% single-walled carbon nanotubes (SW) with carboxyl and amide functionalities. The elemental composition of the epoxy networks was monitored by XPS, showing the increase of O/C ratio to 0.387 when carboxyl-functionalized SW are added. To achieve sustainable composite materials, citric acid was used as curing agent, as a substitute for conventional counterparts. The influence of both surface functional groups and concentration of SW was evaluated through structural and thermo-mechanical analysis. The progressive increase of the DSC enthalpy for SW formulated systems indicates a possible pattern for specific interactions within the bio-based epoxy translated by adjusted activation energy. For 1% neat SW addition, the Ea values decreased to 46 kJ/mol in comparison with 53 kJ/mol calculated for neat epoxy. Furthermore, the -COOH groups from SW nanostructures exerted a strong influence over the mechanical performance of bio-epoxy networks, improving the crosslinking density with ~ 60% and twofold the storage modulus value. Accordingly, by gradual addition of SW-COOH filler within the ECO-based formulations, a very consistent behaviour in seawater was noted, with a 28% decreased value for the absorption degree.Graphical

Dynamic hierarchical intrusion detection task offloading in IoT edge networks

AbstractThe Internet of Things (IoT) has gained widespread importance in recent time. However, the related issues of security and privacy persist in such IoT networks. Owing to device limitations in terms of computational power and storage, standard protection approaches cannot be deployed. In this article, we propose a lightweight distributed intrusion detection system (IDS) framework, called FCAFE‐BNET (Fog based Context Aware Feature Extraction using BranchyNET). The proposed FCAFE‐BNET approach considers versatile network conditions, such as varying bandwidths and data loads, while allocating inference tasks to cloud/edge resources. FCAFE‐BNET is able to adjust to dynamic network conditions. This can be advantageous for applications with particular quality of service requirements, such as video streaming or real‐time communication, ensuring a steady and reliable performance. Early exit deep neural networks (DNNs) have been employed for faster inference generation at the edge. Often, the weights that the model learns in the initial layer may be sufficiently qualified to perform the required classification tasks. Instead of using subsequent layers of DNNs for generating the inference, we have employed the early‐exit mechanism in the DNNs. Such DNNs help to predict a wide range of testing samples through these early‐exit branches, upon crossing a threshold. This method maintains the confidence values corresponding to the inference. Employing this approach, we achieved a faster inference, with significantly high accuracy. Comparative studies exploit manual feature extraction techniques, that can potentially overlook certain valuable patterns, thus degrading classification performance. The proposed framework converts textual/tabular data into 2‐D images, allowing the DNN model to autonomously learns its own features. This conversion scheme facilitated the identification of various intrusion types, ranging from 5 to 14 different categories. FCAFE‐BNET works for both network‐based and host‐based IDS: NIDS and HIDS. Our experiments demonstrate that, in comparison with recent approaches, FCAFE‐BNET achieves a 39.12%–50.23% reduction in the total inference time on benchmark real‐world datasets, such as: NSL‐KDD, UNSW‐NB 15, ToN_IoT, and ADFA_LD.

A Study of Optimization Algorithms for English Vocabulary Memory and Review Plans: Integrating the Principles of Forgetting Curve and Memory Reinforcement

Mastering English vocabulary is a crucial steppingstone towards achieving fluency in the language. Building a strong English vocabulary foundation is essential for spoken and written fluency. This inefficiency can lead to wasted time reviewing known words or forgetting recently learned ones. This work proposes a novel deep learning-based system for personalized English vocabulary learning: the Contextual Awareness Controlled Generative Adversarial Network with Forget Curve &amp; Memory Reinforcement (CACGAN-FCMR-PEVL) model. CACGAN-FCMR-PEVL leverages deep learning to create a personalized optimization algorithm, analyzing user data and incorporating forgetting curve principles. It utilizes a self-attention mechanism and a conditional generative adversarial network (CGAN) to generate personalized vocabulary maps and review schedules, optimizing memory retention through memory reinforcement techniques. Hyperparameter tuning with the Binary Waterwheel Plant Optimization Algorithm (BWpOA) further enhances the model's effectiveness. When compared to other existing methods, the proposed CACGAN-BWpOA-FCMR-PEVL model shows 53.55%, 31.703% and 32.403% higher Vocabulary Recall and 49.46%, 58.06% and 30.98% higher Word Similarity, indicating superior learning and retention. It also attains 47.42%, 56.701% and 73.21% higher Review Interval Prediction Accuracy (RIPA), which suggests effective personalization based on forgetting curves. These findings suggest that CACGAN-BWpOA-FCMR-PEVL effectively personalizes vocabulary learning, leading to improved memory retention.

Revolutionizing power grid loss prediction with advanced hybrid time series deep learning model

Accurate prediction of grid loss in power distribution networks is pivotal for efficient energy management and pricing strategies. Traditional forecasting approaches often struggle to capture the complex temporal dynamics and external influences inherent in grid loss data. In response, this research presents a novel hybrid time-series deep learning model: Gated Recurrent Units with Temporal Convolutional Networks (GRU-TCN), designed to enhance grid loss prediction accuracy. The proposed model integrates the temporal sensitivity of GRU with the local context awareness of TCN, exploiting their complementary strengths. A learnable attention mechanism fuses the outputs of both architectures, enabling the model to discern significant features for accurate prediction. The model is evaluated using well-established metrics across distinct temporal phases: training, testing, and future projection. Results showcase Resulting in encouraging Figures for mean absolute error, root mean squared error, and mean absolute percentage error, the model’s capacity to capture both long-term trends and transitory patterns. The GRU-TCN hybrid model represents a pioneering approach to power grid loss prediction, offering a flexible and precise tool for energy management. This research not only advances predictive accuracy but also lays the foundation for a smarter and more sustainable energy ecosystem, poised to transform the landscape of energy forecasting.

Designing Writing Across the Professions (WAP) programs at the intersection of work-integrated learning and writing transfer research

In our information age, written communication has become increasingly important in many professions. As a result, university faculty and administrators need to develop specific curricula and pedagogies that will facilitate the process of equipping students with the required writing knowledge and skills to meet the demands of their workplace environments. In this article, we argue that Writing Across the Professions (WAP) as a curricular model meets that requirement, particularly in Work-Integrated Learning (WIL) contexts, which we believe are conducive to fostering writing transfer in university students. WAP foregrounds the importance of writing in workplace contexts and aims to facilitate the transfer of students’ knowledge and practices by focusing on rhetorical genre theory and analysis, discourse community theory and analysis, providing engaged feedback on students’ writing, and inviting students to critically reflect on their previous and current writing knowledge and practices. In this article, we propose four conceptual foundations that university faculty and administrators can utilize to implement WAP programs at their institutions. The first concept is that professional (writing) knowledge and practices are contextual and require lifelong learning; WIL faculty and students thus need to be informed about what is involved in learning to write across professions. Secondly, as the transfer of professional (writing) knowledge and practices goes beyond disciplinary boundaries, both faculty and students need to build contextual awareness. Thirdly, as problem-solving is an integral part of writing in the professions, faculty and students need to engage in critical reflection. Finally, professional (writing) knowledge and practices impact identities and therefore require mentoring. In outlining these shared concepts from WIL and writing transfer research, this article offers examples of how they can inform curricular approaches and pedagogical practices in WAP.

Distilling Knowledge from a Transformer-Based Crack Segmentation Model to a Light-Weighted Symmetry Model with Mixed Loss Function for Portable Crack Detection Equipment

The detection of cracks is extremely important for maintenance of concrete structures. Deep learning-based segmentation models have achieved high accuracy in crack segmentation. However, mainstream crack segmentation models have very high computational complexity, and therefore cannot be used in portable crack detection equipment. To address this problem, a knowledge distilling structure is designed by us. In this structure, a large teacher model named TBUNet is proposed to transfer crack knowledge to a student model with symmetry structure named ULNet. In the TBUNet, stacked transformer modules are used to capture dependency relationships between different crack positions in feature maps and achieve contextual awareness. In the ULNet, only a tiny U-Net with light-weighted parameters is used to maintain very low computational complexity. In addition, a mixed loss function is designed to ensure detail and global features extracted by the teacher model are consistent with those of the student model. Our designed experiments demonstrate that the ULNet can achieve accuracies of 96.2%, 87.6%, and 75.3%, and recall of 97.1%, 88.5%, and 76.2% on the Cracktree200, CRACK500, and MICrack datasets, respectively, which is 4–6% higher than most crack segmentation models. However, the ULNet only has a model size of 1 M, which is suitable for use in portable crack detection equipment.