Gliomas, primary brain tumors arising from glial cells, can be effectively identified using Magnetic Resonance Imaging (MRI), a widely employed diagnostic tool in clinical settings. Accurate glioma segmentation, which is crucial for diagnosis and surgical intervention, can be achieved by integrating multiple MRI modalities that offer complementary information. However, limited access to multiple modalities in certain clinical contexts often results in suboptimal performance of glioma segmentation methods. This study introduces a novel generalized knowledge distillation framework designed to transfer multimodal knowledge from a teacher model to a unimodal student model via two distinct distillation strategies: segmentation graph distillation and cascade region attention distillation. The former enables the student to replicate the teacher’s softened output, whereas the latter facilitates extraction and learning of region feature information at various levels within the teacher model. Our evaluation of the proposed distillation strategies using the BraTS 2018 dataset confirms their superior performance in unimodal segmentation contexts compared with existing methods. Full article

With the rapid development of Internet of Things technology, sharing data safely and efficiently in different Internet of Things enterprises is becoming increasingly urgent. Traditional schemes usually use third-party centralized cloud storage and a single central authoritative organization to realize data storage and access management during data sharing. However, this centralized scheme design has the potential for a single point of failure. When the cloud storage platform is subjected to malicious attacks, it may lead to data loss or privacy leakage problems. Secondly, there is a trust crisis in the design of authoritative central organizations, and centralized rights management makes the data sharing process opaque. In order to address these shortcomings, an improved blockchain and elliptic curve-based multi-authority attribute access control scheme is proposed. Firstly, the interplanetary file system is used to store the ciphertext of symmetric encryption data to solve data leakage and tampering in centralized cloud storage. Secondly, the elliptic curve cryptography-based improved multi-authority ciphertext policy attribute-based encryption algorithm is used to encrypt the symmetric key. It can solve the single point of failure problem of user attribute management and significantly reduce the attribute encryption algorithm’s time and resource consumption. Thirdly, the data-related information is uploaded through the smart contract, and the attribute access threshold is set. Only qualified users can view the private information. Finally, the simulation experiments evaluate the efficiency and effectiveness of the scheme from three perspectives: data storage, smart contract, and attribute encryption. Full article

Video saliency object detection is one of the classic research problems in computer vision, yet existing works rarely focus on the impact of input quality on model performance. As optical flow is a key input for video saliency detection models, its quality significantly affects model performance. Traditional optical flow models only calculate the optical flow between two consecutive video frames, ignoring the motion state of objects over a period of time, leading to low-quality optical flow and reduced performance of video saliency object detection models. Therefore, this paper proposes a new optical flow model that improves the quality of optical flow by expanding the flow perception range and uses high-quality optical flow to enhance the performance of video saliency object detection models. Experimental results on the datasets show that the proposed optical flow model can significantly improve optical flow quality, with the S-M values on the DAVSOD dataset increasing by about 39%, 49%, and 44% compared to optical flow models such as PWCNet, SpyNet, and LFNet. In addition, experiments that fine-tuning the benchmark model LIMS demonstrate that improving input quality can further improve model performance. Full article

With the growing prevalence of modern technologies as part of everyday life, mid-air gestures have become a promising input method in the field of human–computer interaction. This paper analyses the gestures of actual users to define a preliminary gesture vocabulary for home air conditioning (AC) systems and suggests a gesture vocabulary for controlling the AC that applies to both home and vehicle environments. In this study, a user elicitation experiment was conducted. A total of 36 participants were filmed while employing their preferred hand gestures to manipulate a home air conditioning system. Comparisons were drawn between our proposed gesture vocabulary (HomeG) and a previously proposed gesture vocabulary which was designed to identify the preferred hand gestures for in-vehicle air conditioners. The findings indicate that HomeG successfully identifies and describes the employed gestures in detail. To gain a gesture taxonomy that is suitable for manipulating the AC at home and in a vehicle, some modifications were applied to HomeG based on suggestions from other studies. The modified gesture vocabulary (CrossG) can identify the gestures of our study, although CrossG has a less detailed gesture pattern. Our results will help designers to understand user preferences and behaviour prior to designing and implementing a gesture-based user interface. Full article

Traditional trajectory-planning methods are unable to achieve time optimization, resulting in slow response times to unexpected situations. To address this issue and improve the smoothness of joint trajectories and the movement time of quadruped robots, we propose a trajectory-planning method based on time optimization. This approach improves the whale optimization algorithm with simulated annealing (IWOA-SA) together with adaptive weights to prevent the whale optimization algorithm (WOA) from falling into local optima and to balance its exploration and exploitation abilities. We also use Markov chains of stochastic process theory to analyze the global convergence of the proposed algorithm. The results show that our optimization algorithm has stronger optimization ability and stability when compared to six representative algorithms using six different test function suites in multiple dimensions. Additionally, the proposed optimization algorithm consistently constrains the angular velocity of each joint within the range of kinematic constraints and reduces joint running time by approximately 6.25%, which indicates the effectiveness of this algorithm. Full article

We aimed to research the design and path-planning methods of an intelligent disinfection-vehicle system. A ROS (robot operating system) system was utilized as the control platform, and SLAM (simultaneous localization and mapping) technology was used to establish an indoor scene map. On this basis, a new path-planning method combining the A* algorithm and the Floyd algorithm is proposed to ensure the safety, efficiency, and stability of the path. Simulation results show that with the average shortest distance between obstacles and paths of 0.463, this algorithm reduces the average numbers of redundant nodes and turns in the path by 70.43% and 31.1%, respectively, compared to the traditional A* algorithm. The algorithm has superior performance in terms of safety distance, path length, and redundant nodes and turns. Additionally, a mask recognition and pedestrian detection algorithm is utilized to ensure public safety. The results of the study indicate that the method has satisfactory performance. The intelligent disinfection-vehicle system operates stably, meets the indoor mapping requirements, and can recognize pedestrians and masks. Full article

At present, compared to 3D convolution, 2D convolution is less computationally expensive and faster in stereo matching methods based on convolution. However, compared to the initial cost volume generated by calculation using a 3D convolution method, the initial cost volume generated by 2D convolution in the relevant layer lacks rich information, resulting in the area affected by illumination in the disparity map having a lower robustness and thus affecting its accuracy. Therefore, to address the lack of rich cost volume information in the 2D convolution method, this paper proposes a multi-scale adaptive cost attention and adaptive fusion stereo matching network (MCAFNet) based on AANet+. Firstly, the extracted features are used for initial cost calculation, and the cost volume is input into the multi-scale adaptive cost attention module to generate attention weight, which is then combined with the initial cost volume to suppress irrelevant information and enrich the cost volume. Secondly, the cost aggregation part of the model is improved. A multi-scale adaptive fusion module is added to improve the fusion efficiency of cross-scale cost aggregation. In the Scene Flow dataset, the EPE is reduced to 0.66. The error matching rates in the KITTI2012 and KITTI2015 datasets are 1.60% and 2.22%, respectively. Full article

Matrix multiplication is ubiquitous in high-performance applications. It will be a significant part of exascale workloads where power is a big concern. This work experimentally studied the power efficiency of three matrix multiplication algorithms: the definition-based, Strassen’s divide-and-conquer, and an optimized divide-and-conquer. The study used reliable on-chip integrated voltage regulators for measuring the power. Interactions with memory, mainly cache misses, were thoroughly investigated. The main result was that the optimized divide-and-conquer algorithm, which is the most time-efficient, was also the most power-efficient, but only for cases that fit in the cache. It consumed drastically less overall energy than the other two methods, regardless of placement in memory. For matrix sizes that caused a spill to the main memory, the definition-based algorithm consumes less power than the divide-and-conquer ones at a high total energy cost. The findings from this study may be of interest when cutting power usage is more vital than running for the shortest possible time or least amount of energy. Full article

Brand identity (BI) refers to the individual characteristics of an enterprise or a certain brand in the market and in the mind of the public. It reflects the evaluation and recognition of the public on the brand and is the core of the market strategy. Successful BI management can bring great business value. Nowadays, the BI management methods based on Internet, big data, and AI are widely adopted. However, they are also confronted with problems, such as accuracy, effectiveness, and sustainability, especially for the Chinese BI. Our work applies the knowledge graph (KG) and never-ending learning (NEL) for exploring efficient Chinese BI management methods. We adapt the NEL framework for the sustainability. In order to improve the accuracy and effectiveness, we express the BI knowledge with KGs and propose two methods in the subsystem components of NEL: (1) the BI evaluation model based on KG and two-dimensional bag-of-words; (2) the Apriori based on KG. In the knowledge integrator of NEL, we propose the synonym KGs for suppressing the concept duplication and drift. The experimental results show that our method reached high consistency with the experts of BI management and the industry reports. Full article

Machine learning (ML)-based network intrusion detection systems (NIDSs) depend entirely on the performance of machine learning models. Therefore, many studies have been conducted to improve the performance of ML models. Nevertheless, relatively few studies have focused on the feature set, which significantly affects the performance of ML models. In addition, features are generated by analyzing data collected after the session ends, which requires a significant amount of memory and a long processing time. To solve this problem, this study presents a new session feature set to improve the existing NIDSs. Current session-feature-based NIDSs are largely classified into NIDSs using a single-host feature set and NIDSs using a multi-host feature set. This research merges two different session feature sets into an integrated feature set, which is used to train an ML model for the NIDS. In addition, an incremental feature generation approach is proposed to eliminate the delay between the session end time and the integrated feature creation time. The improved performance of the NIDS using integrated features was confirmed through experiments. Compared to a NIDS based on ML models using existing single-host feature sets and multi-host feature sets, the NIDS with the proposed integrated feature set improves the detection rate by 4.15% and 5.9% on average, respectively. Full article

In this study, we: (a) compared the differences in the learning of friction concepts between a physical manipulative tool (PMT) and a visuohaptic simulation (VHS) in four different configurations (visually enhanced feedback on/off, force feedback on/off), and (b) analyzed the influence of the visual and haptic feedback for learning the concept of friction. Specifically, this study explored the role of an object’s size in friction. In a three-stage experiment (i.e., pre-test, experimentation, and post-test), 206 undergraduate students compared the friction force, speed, acceleration, and traveled distance between two cubes with the same weight but different sizes pushed on a smooth surface. Our results suggest that (a) VHS was an effective tool for promoting the learning of friction concepts actively, (b) learners in the VHS condition outperformed the learners in the PMT condition (PMT < VHS), (c) the easy identification of the forces by enhanced visual cues promoted the acquisition of scientific knowledge, (d) the haptic feedback promoted a grounded experience for learning about friction, and (e) learners in the Sequenced (H→H + V) condition had more learning benefits than the Simultaneous (H + V), Visual, and Haptic conditions. Students in the Sequenced (H→H + V) condition took advantage of the affordances of the virtual and physical manipulatives. The implication for teaching and learning is that the virtual and physical affordances of the learning tools and the students’ prior knowledge must be considered in the design of the VHS to enhance learning. For the education research, the study implied that body actions positively impacted the learning experience. Full article

As an important topological property for a 3D binary image, the Euler number can be computed by finding specific a voxel block with 2 × 2 × 2 voxels, named the voxel pattern, in the image. In this paper, we introduce three strategies for enhancing the efficiency of a voxel-pattern-based Euler number computing algorithm used for 3D binary images. The first strategy is taking advantage of the voxel information acquired during computation to avoid accessing voxels repeatedly. This can reduce the average number of accessed voxels from 8 to 4 for processing a voxel pattern. Therefore, the efficiency of computation will be improved. The second strategy is scanning every two rows and processing two voxel patterns simultaneously in each scan. In this strategy, only three voxels need to be accessed when a voxel pattern is processed. The last strategy is determining the voxel accessing order in the processing voxel pattern and unifying the processing of the voxel patterns that have identical Euler number increments to one group in the computation. Although this strategy can theoretically reduce the average number of voxels accessed from 8 to 4.25 for processing a voxel pattern, it is more efficient than the above two strategies for moderate- and high-density 3D binary images. Experimental results demonstrated that the three algorithms with each of our proposed three strategies exhibit greater efficiency compared to the conventional Euler number computing algorithm based on finding specific voxel patterns in the image. Full article

As in-vehicle information systems (IVIS) grow increasingly complex, the demand for innovative artificial intelligence-based interaction methods that enhance cybersecurity becomes more crucial. In-air gestures offer a promising solution due to their intuitiveness and individual uniqueness, potentially improving security in human–computer interactions. However, the impact of in-air gestures on driver distraction during in-vehicle tasks and the scarcity of skeleton-based in-air gesture recognition methods in IVIS remain largely unexplored. To address these challenges, we developed a skeleton-based framework specifically tailored for IVIS that recognizes in-air gestures, classifying them as static or dynamic. Our gesture model, tested on the large-scale AUTSL dataset, demonstrates accuracy comparable to state-of-the-art methods and increased efficiency on mobile devices. In comparative experiments between in-air gestures and touch interactions within a driving simulation environment, we established an evaluation system to assess the driver’s attention level during driving. Our findings indicate that in-air gestures provide a more efficient and less distracting interaction solution for IVIS in multi-goal driving environments, significantly improving driving performance by 65%. The proposed framework can serve as a valuable tool for designing future in-air gesture-based interfaces for IVIS, contributing to enhanced cybersecurity. Full article

In recent years, researchers have proposed many deep learning algorithms for data representation learning. However, most deep networks require extensive training data and a lot of training time to obtain good results. In this paper, we propose a novel deep learning method based on stretching deep architectures that are composed of stacked feature learning models. Hence, the method is called “stretching deep architectures” (SDA). In the feedforward propagation of SDA, feature learning models are firstly stacked and learned layer by layer, and then the stretching technique is applied to map the last layer of the features to a high-dimensional space. Since the feature learning models are optimized effectively, and the stretching technique can be easily calculated, the training of SDA is very fast. More importantly, the learning of SDA does not need back-propagation optimization, which is quite different from most of the existing deep learning models. We have tested SDA in visual texture perception, handwritten text recognition, and natural image classification applications. Extensive experiments demonstrate the advantages of SDA over traditional feature learning models and related deep learning models. Full article

The success of businesses is now mostly dependent on e-learning methods as these methods are a rapidly growing innovative technology. Blockchain technology has also been considered to have the ability to change businesses. Therefore, this research aims to explore the direct influence of e-learning on the effectiveness of privacy and security in electronics manufacturing. This study also examines the considerable mediating role of the adoption of blockchain technology between e-learning and privacy and security. Furthermore, the current research investigates how digital orientation moderates the association between e-learning and privacy and security. For the collection of data, the cross-sectional research design and random sampling technique were used, and data were gathered from employees of electronics manufacturing firms in Pakistan through questionnaires. The working response rate of the study was 70%. The findings proved that e-learning plays a considerable role in boosting the privacy and security of electronics manufacturers. The results also demonstrate that the adoption of blockchain technology mediates and digital orientation moderates the link between e-learning and privacy and security. This study adds to the better understanding of management by presenting the significant role of e-learning and blockchain technology in improving the efficiency of privacy and security for electronics manufacturing firms. Full article