Learning Scene Research Articles

Breathing democratic air: North Korean refugees’ journey to sociocultural learning about democracy

This study investigates the acquisition of normative knowledge and values about democracy by North Korean refugees in the United States through everyday experiences and academic concepts. Drawing on Vygotsky’s sociocultural theory, which posits that knowledge is socially constructed and forms the basis of a society’s culture, an analysis is conducted on the sociocultural learning experiences of democracy among North Korean refugees. Utilizing Clarke’s situational analysis of interview transcripts, various scenes of sociocultural learning about democracy and the interpretation of these experiences by North Korean refugees compared to their past in North Korea are uncovered. A common space (e.g. ESL courses) for teaching refugees about democracy can serve as a valuable platform for non-traditional, adult refugee learners, regardless of their educational background, affordability or status. The study’s findings underscore the significance of providing refugees with opportunities for democratic education in a new cultural context, with potential implications for democratic citizenship education.

HyperSOR: Context-Aware Graph Hypernetwork for Salient Object Ranking.

Salient object ranking (SOR) aims to segment salient objects in an image and simultaneously predict their saliency rankings, according to the shifted human attention over different objects. The existing SOR approaches mainly focus on object-based attention, e.g., the semantic and appearance of object. However, we find that the scene context plays a vital role in SOR, in which the saliency ranking of the same object varies a lot at different scenes. In this paper, we thus make the first attempt towards explicitly learning scene context for SOR. Specifically, we establish a large-scale SOR dataset of 24,373 images with rich context annotations, i.e., scene graphs, segmentation, and saliency rankings. Inspired by the data analysis on our dataset, we propose a novel graph hypernetwork, named HyperSOR, for context-aware SOR. In HyperSOR, an initial graph module is developed to segment objects and construct an initial graph by considering both geometry and semantic information. Then, a scene graph generation module with multi-path graph attention mechanism is designed to learn semantic relationships among objects based on the initial graph. Finally, a saliency ranking prediction module dynamically adopts the learned scene context through a novel graph hypernetwork, for inferring the saliency rankings. Experimental results show that our HyperSOR can significantly improve the performance of SOR.

Autistic re-creation: Live action role-play as a safe and brave space for social learning

Live action role-play (larp) is an improvised and embodied form of storytelling that develops through co-creation. While autistic individuals generally dislike unpredictable situations and social interaction in larger groups, anecdotal reports suggest that they often are found in the larping community. Through interviews with 10 autistic, experienced adult larpers, this study investigates larp as a scene for social learning. The examination focuses on the attraction of larps; what the participants learn about social interaction through larping; and how such insights transfer to everyday life. From a cognitive point of view, larping comes through as an optimal autistic space where players feel safe enough to dare push their social boundaries, which conceivably contributes to the formation of more precise social predictions since experimentation and mistakes facilitate embodied learning. From the participants’ point of view, larping improved their understanding of social nuances, which in turn increased social confidence and well-being in everyday life.

Balancing Act: Assessing the Impact of AI Tools on Active Learning in IT Practical Labs

Abstract: In a long time, the integration of Counterfeit Insights (AI) instruments in instructive settings has changed the scene of learning, especially within the field of Information Technology (IT). This term paper digs into the concerning perception of students' expanding dependence on AI instruments amid viable labs, possibly ruining their dynamic engagement and problemsolving abilities. The ponder points to explore the degree of this dependence, its results on students' learning results, and propose rules for teachers to strike a adjust between leveraging AI apparatuses and cultivating dynamic learning. The investigate utilizes a mixed-methods approach, combining a organized study managed to IT und

Learning Scene Representations for Human-assistive Displays Using Self-attention Networks

Video-see-through (VST) augmented reality (AR) is widely used to present novel augmentative visual experiences by processing video frames for viewers. Among VST AR systems, assistive vision displays aim to compensate for low vision or blindness, presenting enhanced visual information to support activities of daily living for the vision impaired/deprived. Despite progress, current assistive displays suffer from a visual information bottleneck, limiting their functional outcomes compared to healthy vision. This motivates the exploration of methods to selectively enhance and augment salient visual information. Traditionally, vision processing pipelines for assistive displays rely on hand-crafted, single-modality filters, lacking adaptability to time-varying and environment-dependent needs. This article proposes the use of Deep Reinforcement Learning (DRL) and Self-attention (SA) networks as a means to learn vision processing pipelines for assistive displays. SA networks selectively attend to task-relevant features, offering a more parameter—and compute-efficient approach to RL-based task learning. We assess the feasibility of using SA networks in a simulation-trained model to generate relevant representations of real-world states for navigation with prosthetic vision displays. We explore two prosthetic vision applications, vision-to-auditory encoding, and retinal prostheses, using simulated phosphene visualisations. This article introduces SA-px, a general-purpose vision processing pipeline using self-attention networks, and SA-phos, a display-specific formulation targeting low-resolution assistive displays. We present novel scene visualisations derived from SA image patches importance rankings to support mobility with prosthetic vision devices. To the best of our knowledge, this is the first application of self-attention networks to the task of learning vision processing pipelines for prosthetic vision or assistive displays.

Language Learning Through AI Technology - LLA (Language Learning Apps)

Learning is the soul of life. Every living being learns new things each and every day. Especially Language learning is a vital role from past to present generation. The modern world needs a skilled and more efficient people for handling different situations. Innovation was the only slogan for language experts.AI technology opens new window for those who are aspiring and developing language learning. Sounds and signs are the basic elements for language components. In the past, language learning was purely transmitted through oral method. After the peak of technological advancement teaching and learning perspective completely changed. The modern world drifted for AI technology enhances and redefined. English language learning scenes in new approaches Like Dulingo, Babbel, Memrise etc. If one is interested in enhancing language proficiency these apps really help and make a great difference in the teaching and learning process.

Learning Generalized Segmentation for Foggy-Scenes by Bi-directional Wavelet Guidance

Learning scene semantics that can be well generalized to foggy conditions is important for safety-crucial applications such as autonomous driving. Existing methods need both annotated clear images and foggy images to train a curriculum domain adaptation model. Unfortunately, these methods can only generalize to the target foggy domain that has seen in the training stage, but the foggy domains vary a lot in both urban-scene styles and fog styles. In this paper, we propose to learn scene segmentation well generalized to foggy-scenes under the domain generalization setting, which does not involve any foggy images in the training stage and can generalize to any arbitrary unseen foggy scenes. We argue that an ideal segmentation model that can be well generalized to foggy-scenes need to simultaneously enhance the content, de-correlate the urban-scene style and de-correlate the fog style. As the content (e.g., scene semantic) rests more in low-frequency features while the style of urban-scene and fog rests more in high-frequency features, we propose a novel bi-directional wavelet guidance (BWG) mechanism to realize the above three objectives in a divide-and-conquer manner. With the aid of Haar wavelet transformation, the low frequency component is concentrated on the content enhancement self-attention, while the high frequency component is shifted to the style and fog self-attention for de-correlation purpose. It is integrated into existing mask-level Transformer segmentation pipelines in a learnable fashion. Large-scale experiments are conducted on four foggy-scene segmentation datasets under a variety of interesting settings. The proposed method significantly outperforms existing directly-supervised, curriculum domain adaptation and domain generalization segmentation methods. Source code is available at https://github.com/BiQiWHU/BWG.

Fair Representation Learning with Maximum Mean Discrepancy Distance Constraint (Student Abstract)

Unsupervised learning methods such as principal component analysis (PCA), t-distributed stochastic neighbor embedding (t-SNE), and autoencoding are regularly used in dimensionality reduction within the statistical learning scene. However, despite a pivot toward fairness and explainability in machine learning over the past few years, there have been few rigorous attempts toward a generalized framework of fair and explainable representation learning. Our paper explores the possibility of such a framework that leverages maximum mean discrepancy to remove information derived from a protected class from generated representations. For the optimization, we introduce a binary search component to optimize the Lagrangian coefficients. We present rigorous mathematical analysis and experimental results of our framework applied to t-SNE.

Creation of Deep Learning Scenarios in the Network Teaching of Physical Education Technical Courses

The network teaching evaluation of sports professional technical courses has positive significance for the sustainable development of education. And how to establish an effective evaluation model is the key part. The research introduces the creation of in-depth learning scenarios (LS) into the network teaching of sports professional technical courses, and then constructs a new network teaching mode of sports professional technical courses. The Particle Swarm Optimization algorithm - Attention- Long Short-Term Memory network (PSO-Attention-LSTM) Chinese Emotion Classification Model (ECM) is constructed to classify the online evaluation text to realize the evaluation of online teaching. This model combines the improved PSO, Attention and LSTM classification models. The optimal number of hidden layer nodes for LSTM model is about 100, and the optimal data size for batch processing is 25. The overall error rate of online teaching teachers of male and female sports professional skill courses is 10.1%, and the overall error rate of online teaching teachers of general and advanced sports professional skill courses is 12.8%. The application effect of the creation of in-depth learning scene in the network teaching of physical education technical courses is shown. When the classification threshold is 0.6 and 0.8 respectively, the AUC of PSO-Attention-LSTM Chinese ECM is 0.821 and 0.809 respectively. The research institute has put forward that the online teaching platform of sports professional technical courses has extremely high practical application value and has been unanimously praised by network users.

MRI evidence for material-specific encoding deficits and mesial-temporal alterations in presurgical frontal lobe epilepsy patients.

Neuroimaging studies reveal frontal lobe (FL) contributions to memory encoding. Accordingly, memory impairments are documented in frontal lobe epilepsy (FLE). Still, little is known about the structural or functional correlates of such impairments. Particularly, material specificity of functional changes in cerebral activity during memory encoding in FLE is unclear. We compared 24 FLE patients (15 right-sided) undergoing presurgical evaluation with 30 healthy controls on a memory fMRI-paradigm of learning scenes, faces, and words followed by an out-of-scanner recognition task as well as regarding their mesial temporal lobe (mTL) volumes. We also addressed effects of FLE lateralization and performance level (normal vs. low). FLE patients had poorer memory performance and larger left hippocampal volumes than controls. Volume increase seemed, however, irrelevant or even dysfunctional for memory performance. Further, functional changes in FLE patients were right-sided for scenes and faces and bilateral for words. In detail, during face encoding, FLE patients had, regardless of their performance level, decreased mTL activation, while during scene and word encoding only low performing FLE patients had decreased mTL along with decreased FL activation. Intact verbal memory performance was associated with higher right frontal activation in FLE patients but not in controls. Pharmacoresistant FLE has a distinct functional and structural impact on the mTL. Effects vary with the encoded material and patients' performance levels. Thus, in addition to the direct effect of the FL, memory impairment in FLE is presumably to a large part due to functional mTL changes triggered by disrupted FL networks. Frontal lobe epilepsy (FLE) patients may suffer from memory impairment. Therefore, we asked patients to perform a memory task while their brain was scanned by MRI in order to investigate possible changes in brain activation during learning. FLE patients showed changes in brain activation during learning and also structural changes in the mesial temporal lobe, which is a brain region especially relevant for learning but not the origin of the seizures in FLE. We conclude that FLE leads to widespread changes that contribute to FLE patients' memory impairment.

ProbIBR: Fast Image-Based Rendering with Learned Probability-Guided Sampling.

We present a general, fast, and practical solution for interpolating novel views of diverse real-world scenes given a sparse set of nearby views. Existing generic novel view synthesis methods rely on time-consuming scene geometry pre-computation or redundant sampling of the entire space for neural volumetric rendering, limiting the overall efficiency. Instead, we incorporate learned MVS priors into the neural volume rendering pipeline while improving the rendering efficiency by reducing sampling points under the guidance of depth probability distributions. Specifically, fewer but important points are sampled under the guidance of depth probability distributions extracted from the learned MVS architecture. Based on the learned probability-guided sampling, we develop a sophisticated neural volume rendering module that effectively integrates source view information with the learned scene structures. We further propose confidence-aware refinement to improve the rendering results in uncertain, occluded, and unreferenced regions. Moreover, we build a four-view camera system for holographic display and provide a real-time version of our framework for free-viewpoint experience, where novel view images of a spatial resolution of 512×512 can be rendered at around 20 fps on a single GTX 3090 GPU. Experiments show that our method achieves 15 to 40 times faster rendering compared to state-of-the-art baselines, with strong generalization capacity and comparable high-quality novel view synthesis performance.

The use of Scene Text Detection methods for news crawl extraction

News crawl extraction is a subset of media monitoring. It is not directly related to the broadcast itself and not spoken by any anchor or guest - therefore its content cannot be inspected via speech-to-text tools. News crawl content is only available as text in video frames so its extraction can be done using Scene Text Detection techniques. This work tests the most used deep learning Scene Text Detection methods (both regression and semantic segmentation approaches) to see if they can be used as a base in news crawl extraction. Our investigation shows that regression based model CTPN performs better for our particular purpose that the semantic segmentation based EAST and CRAFT models.

Research on practical teaching of Ideology and politics in higher education institutions based on VR technology in the information age

Abstract The application of VR technology to the teaching of college Ideology and politics classes not only brings new opportunities for college education innovation but also innovates the Ideology and politics teaching mode. This paper first introduces the principle of image acquisition and the presentation of virtual reality technology. After determining the parameters of left and right viewpoints, the scene images are acquired through graphic rendering technology, and then the final synthesis of left and right images is carried out. Then the determination of the comfort zone of the stereo display is introduced, using the correspondence between stereo depth and γ to determine the comfort zone. Finally, the effect of virtual reality devices in practical teaching of Ideology and politics is analyzed in terms of student performance, students’ perceived usefulness, satisfaction, and classroom acceptance of VR devices. Regarding perceived usefulness, 41% of students strongly agreed, and 30% agreed with the question, “I think the technology provides a better virtual learning scene”. The integration of VR technology into the practical teaching of Ideology and politics in higher education institutions has broken through the traditional teaching mode of Ideology and politics in higher education and improved the experience of Ideology and politics in higher education.

Deep Graph Structural Infomax

In the scene of self-supervised graph learning, Mutual Information (MI) was recently introduced for graph encoding to generate robust node embeddings. A successful representative is Deep Graph Infomax (DGI), which essentially operates on the space of node features but ignores topological structures, and just considers global graph summary. In this paper, we present an effective model called Deep Graph Structural Infomax (DGSI) to learn node representation. We explore to derive the structural mutual information from the perspective of Information Bottleneck (IB), which defines a trade-off between the sufficiency and minimality of representation on the condition of the topological structure preservation. Intuitively, the derived constraints formally maximize the structural mutual information both edge-wise and local neighborhood-wise. Besides, we develop a general framework that incorporates the global representational mutual information, local representational mutual information, and sufficient structural information into the node representation. Essentially, our DGSI extends DGI and could capture more fine-grained semantic information as well as beneficial structural information in a self-supervised manner, thereby improving node representation and further boosting the learning performance. Extensive experiments on different types of datasets demonstrate the effectiveness and superiority of the proposed method.

Emotion Recognition in Learning Scenes Supported by Smart Classroom and Its Application

Emotion Recognition in Learning Scenes Supported by Smart Classroom and Its Application

Lifelong learning with Shared and Private Latent Representations learned through synaptic intelligence

This paper explores a novel lifelong learning method with Shared and Private Latent Representations (SPLR), which are learned through synaptic intelligence. To solve a sequence of tasks, by considering the entire parameter learning trajectory, SPLR can learn task-invariant representation which changes little, and task-specific features that change greatly along the entire parameter updating trajectory. Therefore, in the lifelong learning scenarios, our model can obtain a task-invariant structure shared by all tasks and also contain some private properties that are task-specific to each task. To reduce the parameter quantity, a ℓ1 regularization to promote sparsity is employed in the weights. We use multiple datasets under lifelong learning scenes to verify our SPLR, on these datasets it can get comparable performance compared with existing lifelong learning approaches, and learn a sparse network which means fewer parameters while requiring less model training time.

D→K→I: Data-Knowledge-Driven Group Intelligence Framework for Smart Service in Education Metaverse

Metaverse is the fusion of cyber–physical–social intelligence, and the fusion becomes the core and fundamental property of the metaverse. As an important part of social operationalization, the education domain leads to the birth of the education metaverse. This article answers three basic questions about smart services in the education metaverse: 1) learning scene; 2) technical framework; and 3) initial expansion. Specifically, four key elements constitute the learning scene in the education metaverse: 1) the learner; 2) its time; 3) space; and 4) learning event. In this learning scene, we propose a novel data-knowledge-driven group intelligence framework, aiming to transform data in the education metaverse into knowledge, and intersect and integrate intelligence with knowledge; based on this framework, we apply it to specific services, i.e., transaction and management services. We hope that our work opens the door to research on smart services in the education metaverse and more scholars will work for these challenges.

Automatic classification of interactive texts in online collaborative discussion based on multi-feature fusion

The recognition of learners' speech intention in the online collaborative learning scene is of great significance for exploring the rules of knowledge construction such as knowledge development and emotional communication in the collaborative process. The essence of speech intention recognition is text classification. At present, text classification is mostly based on deep learning model. However, online collaborative discussion has the characteristics of strong contextual semantic relationship and key characteristic words. When only the deep learning method is used for text classification, there may be insufficient acquisition of contextual semantic relations and neglect of key feature words, resulting in a decline in the accuracy of classification results. Therefore, this paper proposes a multi-feature fusion model, which uses BERT to represent the text as a word vector, BiLSTM to extract the context features of the text, CNN to extract the local features of the text, and average pooling model to extract the average representation features of the text. The results show that the overall classification accuracy of the multi-feature fusion model is 84.30%.

Using Active Learning Strategies to Strengthen Cultural and Linguistic Diversity Training in Communication Sciences and Disorders Programs.

Training in cultural competence is an accreditation requirement for graduate programs in Communication Sciences and Disorders (CSD; Council on Academic Accreditation in Audiology and Speech-Language Pathology, 2020). Cultural and linguistic diversity (CLD) instruction within CSD programs and current instructional models may not provide students with effective training in this area (Hammond et al., 2009; Higby et al., 2021; Stockman et al., 2008). In this paper, we present active learning as an instructional approach that may provide students with stronger training in the assessment and treatment of individuals with unfamiliar cultural and linguistic backgrounds. Active learning emphasizes the creation of a supportive classroom environment, teaching skills rather than content, and fostering metacognition in students (Bransford et al., 2000; Gooblar, 2019). We propose a three-part pedagogical model for using active learning strategies to improve clinical training in assessment and treatment of clients with CLD backgrounds. This pedagogical model encourages instructors to set the scene for learning, present a problem to solve, and build in reflection and generalization. Active learning approaches, as described in the model, are ideal for teaching clinical problem-solving across populations while reflecting on one's lived experience and positionality. Sample materials for readers to create their own lesson plans using the model are provided and reviewed.

Boosting Scene Graph Generation with Visual Relation Saliency

The scene graph is a symbolic data structure that comprehensively describes the objects and visual relations in a visual scene, while ignoring the inherent perceptual saliency of each visual relation (i.e., relation saliency). However, humans often quickly allocate attention to important/salient visual relations in a scene. To align with such human perception of a scene, we explicitly model the perceptual saliency of visual relation in scene graph by upgrading each graph edge (i.e., visual relation) with an attribute of relation saliency. We present a new design, named as Saliency-guided Message Passing (SMP), that boosts the generation of such scene graph structure with the guidance from the visual relation saliency. Technically, an object interaction encoder is first utilized to strengthen object relation representations by jointly exploiting the appearance, semantic, and spatial relations in between. A branch is further leveraged to estimate the relation saliency of each visual relation by ordinal regression. Next, conditioned on the object and relation features (coupled with the estimated relation saliency), our SMP enhances scene graph generation by performing message passing over the objects and the most salient relations. Extensive experiments on VG-KR and VG150 datasets demonstrate the superiority of SMP for the scene graph generation. Moreover, we empirically validate the compelling generalizability of the learned scene graphs via SMP on downstream tasks like cross-model retrieval and image captioning.