教学课程资源库 更多>>

关键词： Prototypes   Semantics   Legged locomotion   Electronic mail   Training   Predictive models   Heavily-tailed distribution   Adaptation models   Context modeling   Computer science   Scene graph generation   debasing   concept learning   memory-based learning  

摘要： Existing Unbiased Scene Graph Generation (USGG) methods only focus on addressing the predicate-level imbalance that high-frequency classes dominate predictions of rare ones, while overlooking the concept-level imbalance. Actually, even if predicates themselves are balanced, there is still a significant concept-imbalance within them due to the long-tailed distribution of contexts (i.e., subject-object combinations). This concept-level imbalance poses a more pervasive and challenging issue compared to the predicate-level imbalance since subject-object pairs are inherently complex in combinations. To address the issue, we propose Multi-Concept Learning (MCL), a novel concept-level balanced learning framework orthogonal to existing SGG methods. MCL first quantifies the concept-level imbalance across predicates in terms of different amounts of concepts, representing as multiple concept-prototypes within the same class. Then, to achieve balanced learning across different concepts (i.e., concept-prototypes), we introduce the Concept-based Balanced Memory (CBM), which guides SGG models in generating balanced representations for concept-prototypes. Furthermore, the Concept Regularization (CR) technique is proposed to effectively help models in aligning relation features to their corresponding concept-prototypes, thereby generating concept-level compact and predicate-level distinctive representations for robust relation recognition. Finally, we introduce a novel metric, mean Context Recall (mCR@K), as a complement to mean Recall (mR@K), to evaluate the model's performance across concepts (determined by contexts) within the same predicate. Extensive experiments demonstrate the remarkable efficacy of our model-agnostic strategy in enhancing the performance of benchmark models on both VG-SGG and OI-SGG datasets, leading to new state-of-the-art achievements in two key aspects: predicate-level unbiased relation recognition and concept-level compositional generability. Code is avai

关键词： Anomaly detection   Feature extraction   Protocols   Filling   Vectors   Image edge detection   Graph neural networks   Decoding   Computer science   Deep learning   attributed network   graph convolutional network   reconstruction   mask-filling   neighborhood association  

摘要： As a widely adopted protocol for anomaly detection in attributed networks, reconstruction error prioritizes comprehensive feature extraction to detect anomalies over interrogating the differential representation between normal and abnormal nodes. Intuitively, in attributed networks, normal nodes and their neighbors often exhibit similarities, whereas abnormal nodes demonstrate behaviors distinct from their neighbors. Hence, normal nodes can be accurately represented through their neighbors and effectively reconstructed. As opposed to normal nodes, abnormal nodes represented by their neighbors may be erroneously reconstructed as normal, resulting in increased reconstruction error. Leveraging from this observation, we propose a novel anomaly detection protocol called Neighbor-aware Mask-Filling Anomaly Detection (NMFAD) for attributed networks, aiming to maximize the variability between original and reconstructed features of abnormal nodes filled with information from their neighbors. Specifically, we utilize random-mask on nodes and integrate them into the backbone Graph Neural Networks (GNNs) to map nodes into a latent space. Subsequently, we fill the masked nodes with embeddings from their neighbors and smooth the abnormal nodes closer to the distribution of normal nodes. This optimization improves the likelihood of the decoder to reconstructing abnormal nodes as normal, thereby maximizing the reconstruction error of abnormal nodes. Experimental results demonstrate that, compared to the existing models, NMFAD exhibits superior *** attributed networks.

关键词： extended reality technologies (XRs)   XR camp curriculum design   cognitive development   learning outcomes  

摘要： This study focused on an XR camp designed to integrate multiple subjects into one course, providing students with an immersive, creative, and engaging interdisciplinary learning experience. This study explored the distribution of cognitive development stages among computer science students (CSSs) and engineering students (ESs). Additionally, we compared differences in learning styles, levels of participation and contributions, and learning outcomes between the two groups and verified whether the degree of participation and contributions could predict learning outcomes. The subjects were 54 students (39 = CSSs, 15 = ESs) who participated in the XR camp course from 2020-2021. The students completed the active experiment scale in the 1st week;rated team members using a peer assessment rubric and projects using a work rubric in the 6th, 11th, and 16th weeks;completed the learning ability scales;and participated in interviews in the 18th week. Following their participation in the XR camp, 87% of the students developed high-level cognitive abilities in the XR camp. For the CSSs, course participation positively predicted cognitive development stages, work performance, and collaborative problem-solving abilities, with peer assessments positively influencing work performance. The ESs demonstrated greater interdisciplinary competence, and course participation also positively predicted work performance. These students could also apply XR knowledge and technology to other situations, confirming that the XR camp is conducive to learning for students from diverse backgrounds.

关键词： Prototypes   Semantics   Object detection   Noise   Visualization   Uncertainty   Fuses   Frequency modulation   Computer science   Training   Multimodal fusion   underwater perception   expert learning   uncertainty guidance  

摘要： Existing underwater salient object detection (USOD) methods design fusion strategies to integrate multimodal information, but lack exploration of modal characteristics. To address this, we separately leverage the RGB and depth branches to learn disentangled representations, formulating the heterogeneous experts and hierarchical perception network (HEHP). Specifically, to reduce modal discrepancies, we propose the hierarchical prototype guided interaction (HPI), which achieves fine-grained alignment guided by the semantic prototypes, and then refines with complementary modalities. We further design the mixture of frequency experts (MoFE), where experts focus on modeling high- and low-frequency respectively, collaborating to explicitly obtain hierarchical representations. To efficiently integrate diverse spatial and frequency information, we formulate the four-way fusion experts (FFE), which dynamically selects optimal experts for fusion while being sensitive to scale and orientation. Since depth maps with poor quality inevitably introduce noises, we design the uncertainty injection (UI) to explore high uncertainty regions by establishing pixel-level probability distributions. We further formulate the holistic prototype contrastive (HPC) loss based on semantics and patches to learn compact and general representations across modalities and images. Finally, we employ varying supervision based on branch distinctions to implicitly construct difference modeling. Extensive experiments on two USOD datasets and four relevant underwater scene benchmarks validate the effect of the proposed method, surpassing state-of-the-art binary detection models. Impressive results on seven natural scene benchmarks further demonstrate the scalability.

关键词： Sun   Tensors   Vectors   Optimization   Computer science   Complexity theory   Bipartite graph   Artificial intelligence   Feature extraction   Surges   Dynamic anchor learning   incomplete multi-view clustering   bipartite graph learning   pseudo-label supervise learning   low-rank tensor  

摘要： Compared to static anchor selection, existing dynamic anchor learning could automatically learn more flexible anchors to improve the performance of large-scale multi-view clustering. Despite improving the flexibility of anchors, these methods do not pay sufficient attention to the alignment and fairness of learned anchors. Specifically, within each cluster, the positions and quantities of cross-view anchors may not align, or even anchor absence in some clusters, leading to severe anchor misalignment and imbalance issues. These issues result in inaccurate graph fusion and a reduction in clustering performance. Besides, in practical applications, missing information caused by sensor malfunctions or data losses could further exacerbate anchor misalignment and imbalance. To overcome such challenges, a novel Incomplete Multi-view Clustering with Paired and Balanced Dynamic Anchor Learning (PBDAL) is proposed to ensure the alignment and fairness of anchors. Unlike existing unsupervised anchor learning, we first design a paired and balanced dynamic anchor learning scheme to supervise dynamic anchors to be aligned and fair in each cluster. Meanwhile, we develop an enhanced bipartite graph tensor learning to refine paired and balanced anchors. Our superiority, effectiveness, and efficiency are all validated by performing extensive experiments on multiple public datasets.

关键词： Education   Computer science   Computer science education   Generative adversarial networks   Computational modeling   Indexes   Ethics   Chatbots   Training   Testing   Automated MCQ generation   conditional generative adversarial networks (cGANs)   computer science education   dataset curation   educational assessment  

摘要： This work presents a novel perspective towards generating automated multiple-choice questions (MCQs)-a task fundamentally different due to the highly dynamic nature of computer science education, which spans several sub-domains. Taking advantage of Conditional Generative Adversarial Networks (cGANs), our model provides a versatile approach to addressing the need for diversity and context in relevant MCQ generation across proficiency levels, topic areas. Resulting MCQs inspire implementations within a variety of educational environments - from classrooms, to online courses, and finally exams - equipping teachers with an instrument that could be easily adapted based on the specific needs o students. The model is trained on a carefully constructed dataset that includes material from more than 20 subareas in computer science, consisting of materials such as textbooks, online encyclopedias and Q&A websites. Through rigorous evaluation using comprehensive performance metrics, including Question Relevance Score (QRS), Diversity Index (DI), and Difficulty Alignment Accuracy (DAA), we demonstrate the efficacy and robustness of our framework in generating high-quality MCQs. Moreover, we address ethical considerations inherent in AI-driven educational assessment, ensuring fairness, transparency, and accountability in the MCQ generation process. The cGAN architecture facilitates the generation of contextually relevant MCQs across various proficiency levels and subject domains, enhancing the educational assessment process. The comprehensive dataset developed for this study encompasses diverse computer science topics curated from authoritative textbooks, online resources, question banks, and instructor-generated content. Additionally, a user-friendly QT application has been developed, enabling seamless integration of the cGAN model into educational environments. Through rigorous evaluation and ethical considerations, this framework demonstrates its efficacy, ensuring fairness, tr

关键词： Computer Technology   Railway Transportation   Technological Innovation   Technology Transfer   MPSO Algorithm  

摘要： This article aims to discuss the influence of computer technology on the innovation and technology transfer of modern railway transportation. By historical analysis and empirical research, this article first reviews the evolution of railway transportation from steam locomotive to High-Speed Railways, and then compares and analyzes the performance of the improved multi-objective particle swarm optimization (MPSO) algorithm and traditional particle swarm optimization (PSO) algorithm in train operation scheduling through designing simulation experiments. The experiment selected the representative train operation data sets, and the first 25 and 15 data items are used as training sets respectively to predict the subsequent data items, and the prediction results of MPSO algorithm, traditional PSO algorithm and support vector machine algorithm are compared. The results show that MPSO algorithm is excellent in prediction accuracy, convergence speed and error control, and its prediction results are closer to the actual train running state, which verifies the effectiveness of computer technology in optimizing railway transportation scheduling.

关键词： Feature extraction   Transformers   Convolution   Computational modeling   Data mining   Sun   Mixers   Computer science   Support vector machines   Redundancy   Hyperspectral image classification   hierarchical Transformer   multi-feature   cluster attention  

摘要： Vision Transformer (ViT), known for capturing non-local features, is an effective tool for hyperspectral image classification (HSIC). However, ViT's multi-head self-attention (MHSA) mechanism often struggles to balance local details and long-range relationships for complex high-dimensional data, leading to a loss in spectral-spatial information representation. To address this issue, we propose a deformable convolution-enhanced hierarchical Transformer with spectral-spatial cluster attention (SClusterFormer) for HSIC. The model incorporates a unique cluster attention mechanism that utilizes spectral angle similarity and Euclidean distance metrics to enhance the representation of fine-grained homogenous local details and improve discrimination of non-local structures in 3D HSI and 2D morphological data, respectively. Additionally, a dual-branch multiscale deformable convolution framework augmented with frequency-based spectral attention is designed to capture both the discrepancy patterns in high-frequency and overall trend of the spectral profile in low-frequency. Finally, we utilize a cross-feature pixel-level fusion module for collaborative cross-learning and fusion of the results from the dual-branch framework. Comprehensive experiments conducted on multiple HSIC datasets validate the superiority of our proposed SClusterFormer model, which outperforms existing methods. The source code of SClusterFormer is available at https://***/Fang666666/HSIC_SClusterFormer.

关键词： Software   Codes   Collaboration   Remote working   Project management   Software engineering   Scrum (Software development)   Organizations   Microservice architectures   Computer science   Agile software development   platform engineering   internal development platform   open-source   scrum  

摘要： Platform Engineering (PE) is a growing area of Software Engineering, with many facets, including the elusive concept of Internal Development Platforms (IDPs). They integrate various technologies and tools to support the realization of software projects, especially their managing and developing roles, fostering Agile collaboration and enhancing team productivity. In particular for the new trends of Agile remote working and for Agile applied to the Digital Transformation in the Public Sector. While proprietary IDPs abound, academic research that could naturally lead to open-source IDPs remain scarce. In particular for Scrum, a fundamental Agile methodology in the strategic area of Digital Transformation in the Public Sector, where IDP self-hosting is also appropriate. This paper addresses this gap by proposing the design and realization of the first, to the best of our knowledge, IDP supporting Scrum, being open-source and with self-hosting capabilities. The end-product of our research is the Compositional Agile System (CAS), an extendable microservice architecture-based IDP. Its design, as well as CAS, makes full use of hands-on previous successful experiences of IDP development by members of this team, specifically non-trivial tasks in Italian Public Administration and to foster remote working in University Software Engineering classes, a practice started during the pandemic but now a standard. We offer several architectural and example scenarios where CAS can be of use, concentrating on the Public Administration. Particularly relevant is the use of CAS in analyzing a pillar of Italian Digital Administration available to its citizens, i.e., AppIo. The software is available as an open resource to the research community.

关键词： Computational models   Computer science  

摘要： Automatic disease diagnosis has become increasingly valuable in clinical practice. The advent of large language models (LLMs) has catalyzed a paradigm shift in artificial intelligence, with growing evidence supporting the efficacy of LLMs in diagnostic tasks. Despite the increasing attention in this field, a holistic view is still lacking. Many critical aspects remain unclear, such as the diseases and clinical data to which LLMs have been applied, the LLM techniques employed, and the evaluation methods used. In this article, we perform a comprehensive review of LLM-based methods for disease diagnosis. Our review examines the existing literature across various dimensions, including disease types and associated clinical specialties, clinical data, LLM techniques, and evaluation methods. Additionally, we offer recommendations for applying and evaluating LLMs for diagnostic tasks. Furthermore, we assess the limitations of current research and discuss future directions. To our knowledge, this is the first comprehensive review for LLM-based disease diagnosis.