Existing optimization algorithms are designed based on the finest granularity, i.e., a point. This design limits global search capability and reduces efficiency in global exploration. To solve this problem, we propose a multi-granularity optimization algorithm (MGO), a novel evolutionary algorithm that uses granular-balls instead of points for optimization. MGO distributes a large number of granular-balls across the solution space to explore both critical regions, where optimal solutions are likely located, and non-critical regions for a broader search. Small, fine-granularity granular-balls focus on critical regions, while larger, coarse-granularity ones cover less critical areas. Additionally, there is competition and cooperation among different granular-balls to better utilize computational resources. This multi-granularity representation enhances global search efficiency and accelerates convergence. MGO outperforms many popular state-of-the-art algorithms in the CEC2013 single-objective optimization benchmarks and shows clear advantages in the social network influence maximization (IM) problem. With strong approximation capability, fast convergence speed, and simple design, MGO presents a competitive alternative to most existing optimization algorithms.

The Influence Maximization (IM) problem in social networks has been extensively studied, with greedy algorithms providing accurate and reliable solutions. However, their high computational cost renders them impractical for large-scale networks. Conversely, structure-based heuristic methods reduce computational complexity but often yield suboptimal solutions. To address these challenges, this paper proposes a novel cooperative and competitive multi-granularity fireworks algorithm (CCFWA). Unlike traditional point-based optimization algorithms, CCFWA employs spherical fireworks of varying granularity to enhance the search process. By distributing fireworks of different scales throughout the solution space, the algorithm effectively distinguishes between critical and less significant regions. Fine-granularity fireworks conduct detailed searches in key areas, while coarse-granularity fireworks enable rapid exploration of less important regions. This multi-granularity representation balances global exploration and local exploitation, improving both accuracy and efficiency in solving the IM problem. Experiments on four real-world social networks demonstrate that CCFWA consistently outperforms six state-of-the-art algorithms in terms of computational efficiency and solution quality.

Multi-view online hashing has received much at-tention due to its huge potential in the area of large-scale multimedia retrieval. However, there are still some issues, e.g., how to alleviate the catastrophic forgetting, how to adequately extract high-level semantic information of multi-view streaming data and improve the discrimination of hash models, and how to effectively optimize the binary constraint problem. In this paper, we propose a novel Semantic Consistency based Dual-asymmetric Discrete Online Hashing method, SC-DDOH for short. It adopts a dual-asymmetric distance-based similarity supervision to retain similarities of new data chunk and database. To extract efficient high-level semantic information, an online semantic consistent supervision to mine the semantic related information from word embedding labels. Moreover, an efficient discrete iterative optimization algorithm is introduced to directly learn hash code in the Hamming space. Experiment results on three large-scale multi-view datasets demonstrate the superiority of SC-DDOH over the state-of-the-art baselines.

Traditional change detection methods usually follow the image differencing, change feature extraction, and classification framework, and their performance is limited by such simple image domain differencing and also the hand-crafted features. Recently, the success of deep convolutional neural networks (CNNs) has widely spread across the whole field of computer vision for their powerful representation abilities. Therefore, in this article, we address the remote sensing image change detection problem with deep learning techniques. We first propose an end-to-end dual-branch architecture, termed the W-Net, with each branch taking as input one of the two bitemporal images as in the traditional change detection models. In this way, CNN features with more powerful representative abilities can be obtained to boost the final detection performance. In addition, W-Net performs differencing in the feature domain rather than in the traditional image domain, which greatly alleviates loss of useful information for determining the changes. Furthermore, by reformulating change detection as an image translation problem, we apply the recently popular generative adversarial network (GAN) in which our W-Net serves as the generator, leading to a new GAN architecture for change detection which we call CDGAN. To train our networks and also facilitate future research, we construct a large scale data set by collecting images from Google Earth and provide carefully manually annotated ground truths. Experiments show that our proposed methods can provide fine-grained change detection results superior to the existing state-of-the-art baselines.

In this letter, we address the problem of change detection for remote sensing images from the perspective of visual saliency computation. The proposed method incorporates low-rank-based saliency computation and deep feature representation. First, multilevel convolutional neural network (CNN) features are extracted for superpixels generated using SLIC, in which a fixed-size CNN feature can be formed to represent each superpixel. Then, low-rank decomposition is applied to the change features of the two input images to generate saliency maps that indicate change probabilities of each pixel. Finally, binarized change map can be obtained with a simple threshold. To deal with scale variations, a multiscale fusion strategy is employed to produce more reliable detection results. Extensive experiments on Google Earth and GF-2 images demonstrate the feasibility and effectiveness of the proposed method.

This paper presents a novel approach for urban change detection of high resolution (HR) remote sensing images. To overcome deficiency of traditional pixel-based methods and better annotate HR images, object-based strategies are adopted. Firstly change vector analysis (CVA) and local binary patterns (LBP) are utilized to extract the object-specific features based on the image-objects acquired by multitemporal segmentation. Then sparse representation is further exploited to characterize highly effective sparse features. Finally, the final change map is obtained by support vector machine (SVM) with the pseudotraining set acquired by expectation maximization (EM). Comparative experiments demonstrate the effectiveness of the proposed method.

This paper introduces a mechanism in a ubiquitous learning log system, which is designed to guide learners to participate in the learning activities recommended by the system. In order to provoke learners’ interests on the knowledge, the recommended knowledge is related to both what learners are studying and the learners’ current learning contexts. And to help learners grasp the knowledge, in the second step the mechanism provides learners with relevant learning activities to guide learners to use the recommended knowledge.

In our previous works, we developed a system named SCROLL in order to log, organize, recall and evaluate the learning log. However up to now, we just use an active mode to record logs. This means that a learner must take a capture of learned contents consciously and most of learning chances be lost unconsciously. In order to solve this problem, we started a project named PACALL (Passive Capture for Learning Log) in order to have a passive capture using SenseCam. With the help of SenseCam, learner’s activity can be captured as a series of images. We also developed a system to help a learner find the important images by analyzing sensor data and images processing technology. Finally, the selected images will be uploaded to the current SCROLL system as ubiquitous learning logs. This research suggests that SenseCam can be used to do passive capture of learning experiences and workload of reflection can be reduced by analyzing sensor data of SenseCam.

This study primarily exploits a context-awareness and personalization model supporting ubiquitous learning log system. Learning log stands for the log of knowledge or learning experience acquired ubiquitously. The model has three main behaviors, which are to recommend learning objects in accordance with both learners’ needs and contexts, to detect their learning styles using the context history and to prompt them to review what they have learned regarding their learning styles. What’s more, by monitoring learners’ reaction on the recommendation or prompting, the model can improve its prediction.

This paper proposes a ubiquitous learning log system called SCROLL (System for Capturing and Reminding Of Learning Log). Ubiquitous Learning Log (ULL) is defined as a digital record of what you have learned in the daily life using ubiquitous technologies. It allows you to log your learning experiences with photos, audios, videos, location, QR-code, RFID tag, and sensor data, and to share and to reuse ULL with others. Using SCROLL, you can receive personalized quizzes and answers for your questions. Also, you can navigate and be aware of your past ULLs supported by augmented reality view. The initial evaluation of applying this system in an undergraduate English course is illustrated.