Search Results (17)

Weakly supervised video anomaly detection (WSVAD) aims to identify unusual events using only video-level labels. However, current methods face several key challenges, including ineffective modelling of complex temporal dependencies, indistinct feature boundaries between visually similar normal and abnormal events, and high false alarm rates caused by an inability to distinguish salient events from complex background noise. This paper proposes a novel method that systematically enhances feature representation and discrimination to address these challenges. The proposed method first builds robust temporal representations by employing a hierarchical multi-scale temporal encoder and a position-aware global relation network to capture both local and long-range dependencies. The core of this method is the dual-stream attention-enhanced memory network, which achieves precise discrimination by learning distinct normal and abnormal patterns via dual memory banks, while utilising bidirectional spatial attention to mitigate background noise and focus on salient events before memory querying. The models underwent a comprehensive evaluation utilising solely RGB features on two demanding public datasets, UCF-Crime and XD-Violence. The experimental findings indicate that the proposed method attains state-of-the-art performance, achieving 87.43% AUC on UCF-Crime and 85.51% AP on XD-Violence. This result demonstrates that the proposed “attention-guided prototype matching” paradigm effectively resolves the aforementioned challenges, enabling robust and precise anomaly detection. Full article

This advanced exploration of integrating cross-modal Artificial-Intelligence-Generated Content (AIGC) within the Unity3D game engine seeks to elevate the diversity and coherence of image generation in game art creation. The theoretical framework proposed dives into the seamless incorporation of generated visuals within Unity3D, introducing a novel Generative Adversarial Network (GAN) structure. In this architecture, both the Generator and Discriminator embrace a Transformer model, adeptly managing sequential data and long-range dependencies. Furthermore, the introduction of a cross-modal attention module enables the dynamic calculation of attention weights between text descriptors and generated imagery, allowing for real-time modulation of modal inputs, ultimately refining the quality and variety of generated visuals. The experimental results show outstanding performance on technical benchmarks, with an inception score reaching 8.95 and a Frechet Inception Distance plummeting to 20.1, signifying exceptional diversity and image quality. Surveys reveal that users rated the model’s output highly, citing both its adherence to text prompts and its strong visual allure. Moreover, the model demonstrates impressive stylistic variety, producing imagery with intricate and varied aesthetics. Though training demands are extended, the payoff in quality and diversity holds substantial practical value. This method exhibits substantial transformative potential in Unity3D development, simultaneously improving development efficiency and optimizing the visual fidelity of game assets. Full article

Traditional disease detection methods typically depend on visual assessments conducted by human experts, which are time-consuming and subjective. Thus, there is an urgent demand for automated and efficient approaches to accurately detect and classify tea diseases. This study presents an enhanced Real-Time Detection Transformer (RT-DETR), tailored for the accurate and efficient identification of tea diseases in natural environments. The proposed method integrates three novel components: Faster-LTNet, CG Attention Module, and RMT Spatial Prior Block, to significantly improve computational efficiency, feature representation, and detection capabilities. Faster-LTNet employs partial convolution and hierarchical design to optimize computational resources, while the CG Attention Module enhances multi-head self-attention by introducing grouped feature inputs and cascading operations to reduce redundancy and increase attention diversity. The RMT Spatial Prior Block integrates a Manhattan distance-based spatial decay matrix and linear decomposition strategy to improve global and local context modeling, reducing attention complexity. The enhanced RT-DETR model achieves a detection precision of 89.20% and a processing speed of 346.40 FPS. While the precision improves, the FPS value also increases by 109, which is superior to the traditional model in terms of precision and real-time processing. Additionally, compared to the baseline model, the FLOPs are reduced by 50%, and the overall model size and parameter size are decreased by approximately 50%. These findings indicate that the proposed algorithm is well-suited for efficient, real-time, and lightweight agricultural disease detection. Full article

The attentional control settings (ACSs) can help us efficiently select targets in complex real-world environments. Previous research has shown that category-specific ACS demands more attentional resources than feature-specific ACS. However, comparing natural or alphanumeric categories with color features does not distinguish the effects of processing hierarchy and target-defining properties. The present study employed a spatial cueing paradigm to better understand the effects of target-defining properties and search mode on attentional resources in visual search. The target was defined as a combination of shape feature (shape “X”) and color category (green in different shades), which generated shape-specific ACS (sACS) and color-specific ACS (cACS). The degrees of shape matching (SM), color matching (CM), and spatial validity between the cue and target were manipulated. Search modes were manipulated by changing the homogeneity of distractors in either shape or color dimensions. Results show a main effect of CM across all four experiments, indicating that category can tune on attentional capture consistently. Importantly, the analysis between four experiments found different interactions across experiments, suggesting that the singleton search mode can reduce the effects of ACS and increase the interactions with other factors. In conclusion, this study suggests that the effects of ACS on attentional capture are determined by both target-defining properties and search mode, rather than processing hierarchy. The results indicate that attentional processes are highly dynamic and context-dependent, requiring a flexible allocation of resources to effectively prioritize relevant information. Full article

Demand and market value for pennywort largely depend on the quality of the leaves, which can be affected by various ambient environment or fertigation variables during cultivation. Although early detection of defects in pennywort leaves would enable growers to take quick action, conventional manual detection is laborious and time consuming as well as subjective. Therefore, the objective of this study was to develop an automatic leaf defect detection algorithm for pennywort plants grown under controlled environment conditions, using machine vision and deep learning techniques. Leaf images were captured from pennywort plants grown in an ebb-and-flow hydroponic system under fluorescent light conditions in a controlled plant factory environment. Physically or biologically damaged leaves (e.g., curled, creased, discolored, misshapen, or brown spotted) were classified as defective leaves. Images were annotated using an online tool, and Mask R-CNN models were implemented with the integrated attention mechanisms, convolutional block attention module (CBAM) and coordinate attention (CA) and compared for improved image feature extraction. Transfer learning was employed to train the model with a smaller dataset, effectively reducing processing time. The improved models demonstrated significant advancements in accuracy and precision, with the CA-augmented model achieving the highest metrics, including a mean average precision (mAP) of 0.931 and an accuracy of 0.937. These enhancements enabled more precise localization and classification of leaf defects, outperforming the baseline Mask R-CNN model in complex visual recognition tasks. The final model was robust, effectively distinguishing defective leaves in challenging scenarios, making it highly suitable for applications in precision agriculture. Future research can build on this modeling framework, exploring additional variables to identify specific leaf abnormalities at earlier growth stages, which is crucial for production quality assurance. Full article

In the rapidly evolving online marketplace, accurately predicting the demand for pre-owned items presents a significant challenge for sellers, impacting pricing strategies, product presentation, and marketing investments. Traditional demand prediction methods, while foundational, often fall short in addressing the dynamic and heterogeneous nature of e-commerce data, which encompasses textual descriptions, visual elements, geographic contexts, and temporal dynamics. This paper introduces a novel approach utilizing the Graph Neural Network (GNN) to enhance demand prediction accuracy by leveraging the spatial relationships inherent in online sales data, named SGNN. Drawing from the rich dataset provided in the fourth Kaggle competition, we construct a spatially aware graph representation of the marketplace, integrating advanced attention mechanisms to refine predictive accuracy. Our methodology defines the product demand prediction problem as a regression task on an attributed graph, capturing both local and global spatial dependencies that are fundamental to accurate predicting. Through attention-aware message propagation and node-level demand prediction, our model effectively addresses the multifaceted challenges of e-commerce demand prediction, demonstrating superior performance over traditional statistical methods, machine learning techniques, and even deep learning models. The experimental findings validate the effectiveness of our GNN-based approach, offering actionable insights for sellers navigating the complexities of the online marketplace. This research not only contributes to the academic discourse on e-commerce demand prediction but also provides a scalable and adaptable framework for future applications, paving the way for more informed and effective online sales strategies. Full article

Automatically segmenting polyps from colonoscopy videos is crucial for developing computer-assisted diagnostic systems for colorectal cancer. Existing automatic polyp segmentation methods often struggle to fulfill the real-time demands of clinical applications due to their substantial parameter count and computational load, especially those based on Transformer architectures. To tackle these challenges, a novel lightweight long-range context fusion network, named LightCF-Net, is proposed in this paper. This network attempts to model long-range spatial dependencies while maintaining real-time performance, to better distinguish polyps from background noise and thus improve segmentation accuracy. A novel Fusion Attention Encoder (FAEncoder) is designed in the proposed network, which integrates Large Kernel Attention (LKA) and channel attention mechanisms to extract deep representational features of polyps and unearth long-range dependencies. Furthermore, a newly designed Visual Attention Mamba module (VAM) is added to the skip connections, modeling long-range context dependencies in the encoder-extracted features and reducing background noise interference through the attention mechanism. Finally, a Pyramid Split Attention module (PSA) is used in the bottleneck layer to extract richer multi-scale contextual features. The proposed method was thoroughly evaluated on four renowned polyp segmentation datasets: Kvasir-SEG, CVC-ClinicDB, BKAI-IGH, and ETIS. Experimental findings demonstrate that the proposed method delivers higher segmentation accuracy in less time, consistently outperforming the most advanced lightweight polyp segmentation networks. Full article

For small target objects on fast-moving conveyor belts, traditional vision detection algorithms equipped with conventional robotic arms struggle to capture the long and short-range pixel dependencies crucial for accurate detection. This leads to high miss rates and low precision. In this study, we integrate the traditional EMA (efficient multi-scale attention) algorithm with the c2f (channel-to-pixel) module from the original YOLOv8, alongside a Faster-Net module designed based on partial convolution concepts. This fusion results in the Faster-EMA-Net module, which greatly enhances the ability of the algorithm and robotic technologies to extract pixel dependencies for small targets, and improves perception of dynamic small target objects. Furthermore, by incorporating a small target semantic information enhancement layer into the multiscale feature fusion network, we aim to extract more expressive features for small targets, thereby boosting detection accuracy. We also address issues with training time and subpar performance on small targets in the original YOLOv8 algorithm by improving the loss function. Through experiments, we demonstrate that our attention-based visual detection algorithm effectively enhances accuracy and recall rates for fast-moving small targets, meeting the demands of real industrial scenarios. Our approach to target detection using industrial robotic arms is both practical and cutting-edge. Full article

Theories of holistic face processing vary widely with respect to conceptualizations, paradigms, and stimuli. These divergences have left several theoretical questions unresolved. Namely, the role of attention in face perception is understudied. To rectify this gap in the literature, we combined the complete composite face task (allowing for predictions of multiple theoretical conceptualizations and connecting with a large body of research) with a secondary auditory discrimination task at encoding (to avoid a visual perceptual bottleneck). Participants studied upright, intact faces within a continuous recognition paradigm, which intermixes study and test trials at multiple retention intervals. Within subjects, participants studied faces under full or divided attention. Test faces varied with respect to alignment, congruence, and retention intervals. Overall, we observed the predicted beneficial outcomes of holistic processing (e.g., higher discriminability for Congruent, Aligned faces relative to Congruent, Misaligned faces) that persisted across retention intervals and attention. However, we did not observe the predicted detrimental outcomes of holistic processing (e.g., higher discriminability for Incongruent, Misaligned faces relative to Incongruent, Aligned faces). Because the continuous recognition paradigm exerts particularly strong demands on attention, we interpret these findings through the lens of resource dependency and domain specificity. Full article

When multiple stimuli are simultaneously presented, they compete against each other to be represented in the capacity-limited visual system. This competition increases as stimulus heterogeneity increases. Given that selective attention is a way to resolve this competition, it has been known that the effect of attention on task performance is magnified as the level of competition increases due to increased stimulus heterogeneity. While previous studies showed that stimulus heterogeneity in a task-irrelevant dimension affects task performance, it remains unknown how this kind of stimulus heterogeneity interacts with visual attention and stimulus-driven competition. Here, we found that the process of searching for a target stimulus among non-targets became inefficient as stimulus heterogeneity in a task-irrelevant dimension increased. The results also showed that the magnitude of the attentional cuing effect could be affected by increased heterogeneity. However, this modulation was dependent on the type of varied feature or task demand. We suggest that increased stimulus heterogeneity in a task-irrelevant dimension would increase stimulus-driven competition, which impoverishes the quality of stimulus representations. Full article

This paper applies a new artificial intelligence architecture, the temporal fusion transformer (TFT), for the joint GDP forecasting of 25 OECD countries at different time horizons. This new attention-based architecture offers significant advantages over other deep learning methods. First, results are interpretable since the impact of each explanatory variable on each forecast can be calculated. Second, it allows for visualizing persistent temporal patterns and identifying significant events and different regimes. Third, it provides quantile regressions and permits training the model on multiple time series from different distributions. Results suggest that TFTs outperform regression models, especially in periods of turbulence such as the COVID-19 shock. Interesting economic interpretations are obtained depending on whether the country has domestic demand-led or export-led growth. In essence, TFT is revealed as a new tool that artificial intelligence provides to economists and policy makers, with enormous prospects for the future. Full article

Objectives: In this study, based on the known vestibulo-hippocampal connections, we asked whether mild chronic vestibulopathy leads only to vestibular-related deficits or whether there are effects on hippocampal function, structure, and cognition in general. In more detail, we assessed whether chronic vestibulopathy leads to (a) deficits in vestibular tasks without cognitive demand (balancing), (b) deficits in spatial cognitive tasks that require vestibular input (path integration, rotational memory), (c) deficits in spatial cognitive tasks that do not rely on vestibular input, (d) deficits in general cognitive function, and (e) atrophy in the brain. Methods: A total of 15 patients with chronic uni- or bilateral vestibulopathy (56.8 ± 10.1 years; 4 females) were included in this study and were age- and gender-matched by the control participants (57.6 ± 10.5) in a pairwise manner. Given their clinical symptoms and their deficits of the vestibulo-ocular reflex (VOR) the patients could be classified as being mildly affected. All participants of the underwent the following tests: clinical balance (CBT), triangle completion (TCT) for path integration, rotational memory (RM), the visuo-spatial subset of the Berlin intelligence structure test (BIS-4) and d2-R for attention and concentration, and a structural MRI for gray matter analysis using voxel-based morphometry (VBM). Results: Compared to the healthy controls, the vestibulopathy patients performed significantly worse in terms of CBT, TCT, and RM but showed no differences in terms of the BIS-4 and d2-R. There were also no significant volumetric gray matter differences between the two groups. Conclusions: This study provides evidence that both non-cognitive and cognitive functions that rely on vestibular input (balancing, path integration, rotational memory) are impaired, even in mild chronic vestibulopathy, while other cognitive functions, which rely on visual input (visuo-spatial memory, attention), are unimpaired in this condition, together with an overall intact brain structure. These findings may reflect a segregation between vestibular- and visual-dependent processes in the medial temporal lobe on the one hand and a structure–function dissociation on the other. Full article

As a quasi-public product, fishery insurance has become an important starting point for the construction of the modern fishery industry chain, supply chain and value chain risk management mechanism. We used visual data processing methods and text mining technology to screen policy samples. We then built a fishery insurance policy evaluation system based on the Policy Modeling Consistency (PMC) index model. We combined the PMC index score and PMC surface to quantitatively analyze the policy samples. This paper has four important findings: (1) After three adjustments and developments, the fishery insurance policy has grown in terms of initial attention, changes, and development and gradually matured. (2) A gap exists between the content of the fishing insurance policy text and the actual demand. The scoring results of the policy samples are concentrated in the acceptable range, the policy effects are not satisfactory, and the formulation of fishery insurance policies has weak links that need to be improved. (3) The consistency and effectiveness of fishery insurance policies have developed simultaneously with fishery insurance research, and the practical effects of high-quality fishery insurance policies are conducive to the development of theoretical research. (4) The policy text of fishery insurance has major problems, such as missing joint force of issuing institutions, low professionalism of the text, inadequate subdivision guidance of fishery insurance, weak social effectiveness, high dependence on financial subsidies, lack of incentive sustainability and corresponding laws and regulations and reduction in policy feasibility among others. Considering the above issues, this paper puts forward relevant policy optimization paths and safeguard measures on the basis of giving priority to greater absolute value. Full article

Visual working memory (VWM) resources have been shown to be flexibly distributed according to item priority. This flexible allocation of resources may depend on attentional control, an executive function known to decline with age. In this study, we sought to determine how age differences in attentional control affect VWM performance when attention is flexibly allocated amongst targets of varying priority. Participants performed a delayed-recall task wherein item priority was varied. Error was modelled using a three-component mixture model to probe different aspects of performance (precision, guess-rate, and non-target errors). The flexible resource model offered a good fit to the data from both age groups, but older adults showed consistently lower precision and higher guess rates. Importantly, when demands on flexible resource allocation were highest, older adults showed more non-target errors, often swapping in the item that had a higher priority at encoding. Taken together, these results suggest that the ability to flexibly allocate attention in VWM is largely maintained with age, but older adults are less precise overall and sometimes swap in salient, but no longer relevant, items possibly due to their lessened ability to inhibit previously attended information. Full article

We move our eyes roughly three times every second while searching complex scenes, but covert attention helps to guide where we allocate those overt fixations. Covert attention may be allocated reflexively or voluntarily, and speeds the rate of information processing at the attended location. Reducing access to covert attention hinders performance, but it is not known to what degree the locus of covert attention is tied to the current gaze position. We compared visual search performance in a traditional gaze-contingent display, with a second task where a similarly sized contingent window is controlled with a mouse, allowing a covert aperture to be controlled independently by overt gaze. Larger apertures improved performance for both the mouse- and gaze-contingent trials, suggesting that covert attention was beneficial regardless of control type. We also found evidence that participants used the mouse-controlled aperture somewhat independently of gaze position, suggesting that participants attempted to untether their covert and overt attention when possible. This untethering manipulation, however, resulted in an overall cost to search performance, a result at odds with previous results in a change blindness paradigm. Untethering covert and overt attention may therefore have costs or benefits depending on the task demands in each case. Full article