Search Results (12)

Cross-Domain Few-Shot Learning (CD-FSL) aims to empower machines with the capability to rapidly acquire new concepts across domains using an extremely limited number of training samples from the target domain. This ability hinges on the model’s capacity to extract and transfer generalizable knowledge from a source training set. Studies have indicated that the similarity between source and target-data distributions, as well as the difficulty of target tasks, determine the classification performance of the model. However, the current lack of quantitative metrics hampers researchers’ ability to devise appropriate learning strategies, leading to a fragmented understanding of the field. To address this issue, we propose quantitative metrics of domain distance and target difficulty, which allow us to categorize target tasks into three regions on a two-dimensional plane: near-domain tasks, far-domain low-difficulty tasks, and far-domain high-difficulty tasks. For datasets in different regions, we propose a Divide-and-Conquer Strategy (DCS) to tackle few-shot classification across various target datasets. Empirical results across 15 target datasets demonstrate the compatibility and effectiveness of our approach, improving the model performance. We conclude that the proposed metrics are reliable and the Divide-and-Conquer Strategy is effective, offering valuable insights and serving as a reference for future research on CD-FSL. Full article

Transformers are fundamental and among the most expensive electrical devices in any power transmission and distribution system. Therefore, it is essential to implement powerful maintenance methods to monitor and predict their condition. Due to its many advantages—such as early detection, accurate diagnosis, cost reduction, and rapid response time—dissolved gas analysis (DGA) is regarded as one of the most effective ways to assess a transformer’s condition. In this contribution, we propose a new probabilistic hierarchical intelligent system consisting of five subnetworks of the radial basis functions (RBF) type. Indeed, hierarchical classification minimizes the complexity of the discrimination task by employing a divide-and-conquer strategy, effectively addressing the issue of unbalanced data (a significant disparity between the categories to be predicted). This approach contributes to a more precise and sophisticated diagnosis of transformers. The first subnetwork detects the presence or absence of defects, separating defective samples from healthy ones. The second subnetwork further classifies the defective samples into three categories: electrical, thermal, and cellulosic decomposition. The samples in these categories are then precisely assigned to their respective subcategories by the third, fourth, and fifth subnetworks. To optimize the hyperparameters of the five models, the Linde–Buzo–Gray algorithm is implemented to reduce the number of centers (radial functions) in each subnetwork. Subsequently, a single-layer perceptron is trained to determine the optimal synaptic weights, which connect the intermediate layer to the output layer. The results obtained with our proposed system surpass those achieved with another implemented alternative (a single RBF), with an average sensitivity percentage as high as 96.85%. This superiority is validated by a Student’s t-test, showing a significant difference greater than 5% (p-value < 0.001). These findings demonstrate and highlight the relevance of the proposed hierarchical configuration. Full article

The leakage signals, including electromagnetic, energy, time, and temperature, generated during the operation of password devices contain highly correlated key information, which leads to security vulnerabilities. In traditional encryption algorithms, the length of the key greatly affects the upper limit of its security against cracking. Regarding side-channel attacks on long-key algorithms, traditional template attack methods characterize the energy traces using multivariate Gaussian distribution during the template construction phase. The exhaustive key-guessing process is expected to consume a significant amount of time and computational resources. Therefore, to analyze the effectiveness of obtaining key values from the side information of password devices, we propose an innovative attack method based on a divide-and-conquer logical structure, targeting semi-bytes. We construct a collection of key classification submodules with symmetric correlations. By integrating a differential network model for byte-block sets and an end-to-end direct attack method, we form a holistic symmetric decision framework and propose a key classification structure based on deep transfer learning. This structure consists of three main parts: side information data acquisition, analysis of key-value effectiveness, and determination of attack positions. It employs multiple parallel symmetric subnetworks, effectively improving attack efficiency and reducing the key enumeration range. Experimental results show that the optimal attack accuracy of the network model can reach 91%, with an average attack accuracy of 78%. It overcomes overfitting issues under small sample dataset conditions. Full article

Deep learning has been applied to structural damage detection and achieved great success in recent years, such as the popular structural damage detection methods based on structural vibration response and convolutional neural networks (CNN). However, due to the limited number of vibration response samples that can be acquired in practice for damage detection, the CNN-based models may not be fully trained; thus, their performance for identifying different damage severity as well as the damage locations may be reduced. To solve this issue, in this paper, we follow the strategy of "divide-and-conquer" and propose a two-stage structural damage detection method. Specifically, in the first stage, a 1D-CNN model is constructed to extract the damage features automatically and identify the damage locations. In the second stage, a support vector machine (SVM) model and wavelet packet decomposition technique are combined to further quantify the damage. Experiments are conducted on an eight-level steel frame structure, and the accuracy of the experimental results is greater than 99%, which demonstrates the superiority of the proposed method compared to the state-of-the-art approaches. Full article

Compressed sensing (CS) has been widely used in hyperspectral (HS) imaging to obtain hyperspectral data at a sub-Nyquist sampling rate, lifting the efficiency of data acquisition. Yet, reconstructing the acquired HS data via iterative algorithms is time consuming, which hinders the real-time application of compressed HS imaging. To alleviate this problem, this paper makes the first attempt to adopt convolutional neural networks (CNNs) to reconstruct three-dimensional compressed HS data by backtracking the entire imaging process, leading to a simple yet effective network, dubbed the backtracking reconstruction network (BTR-Net). Concretely, we leverage the divide-and-conquer method to divide the imaging process based on coded aperture tunable filter (CATF) spectral imager into steps, and build a subnetwork for each step to specialize in its reverse process. Consequently, BTR-Net introduces multiple built-in networks which performs spatial initialization, spatial enhancement, spectral initialization and spatial–spectral enhancement in an independent and sequential manner. Extensive experiments show that BTR-Net can reconstruct compressed HS data quickly and accurately, which outperforms leading iterative algorithms both quantitatively and visually, while having superior resistance to noise. Full article

Molecular modeling is widely utilized in subjects including but not limited to physics, chemistry, biology, materials science and engineering. Impressive progress has been made in development of theories, algorithms and software packages. To divide and conquer, and to cache intermediate results have been long standing principles in development of algorithms. Not surprisingly, most important methodological advancements in more than half century of molecular modeling are various implementations of these two fundamental principles. In the mainstream classical computational molecular science, tremendous efforts have been invested on two lines of algorithm development. The first is coarse graining, which is to represent multiple basic particles in higher resolution modeling as a single larger and softer particle in lower resolution counterpart, with resulting force fields of partial transferability at the expense of some information loss. The second is enhanced sampling, which realizes “dividing and conquering” and/or “caching” in configurational space with focus either on reaction coordinates and collective variables as in metadynamics and related algorithms, or on the transition matrix and state discretization as in Markov state models. For this line of algorithms, spatial resolution is maintained but results are not transferable. Deep learning has been utilized to realize more efficient and accurate ways of “dividing and conquering” and “caching” along these two lines of algorithmic research. We proposed and demonstrated the local free energy landscape approach, a new framework for classical computational molecular science. This framework is based on a third class of algorithm that facilitates molecular modeling through partially transferable in resolution “caching” of distributions for local clusters of molecular degrees of freedom. Differences, connections and potential interactions among these three algorithmic directions are discussed, with the hope to stimulate development of more elegant, efficient and reliable formulations and algorithms for “dividing and conquering” and “caching” in complex molecular systems. Full article

Unscented particle filter (UPF) struggles to completely cover the target state space when handling the maneuvering target tracing problem, and the tracking performance can be affected by the low sample diversity and algorithm redundancy. In order to solve this problem, the method of divide-and-conquer sampling is applied to the UPF tracking algorithm. By decomposing the state space, the descending dimension processing of the target maneuver is realized. When dealing with the maneuvering target, particles are sampled separately in each subspace, which directly prevents particles from degeneracy. Experiments and a comparative analysis were carried out to comprehensively analyze the performance of the divide-and-conquer sampling unscented particle filter (DCS-UPF). The simulation result demonstrates that the proposed algorithm can improve the diversity of particles and obtain higher tracking accuracy in less time than the particle swarm algorithm and intelligent adaptive filtering algorithm. This algorithm can be used in complex maneuvering conditions. Full article

Generative adversarial networks (GANs) have a revolutionary influence on sample generation. Maximum mean discrepancy GANs (MMD-GANs) own competitive performance when compared with other GANs. However, the loss function of MMD-GANs is an empirical estimate of maximum mean discrepancy (MMD) and not precise in measuring the distance between sample distributions, which inhibits MMD-GANs training. We propose an efficient divide-and-conquer model, called DC-MMD-GANs, which constrains the loss function of MMD to tight bound on the deviation between empirical estimate and expected value of MMD and accelerates the training process. DC-MMD-GANs contain a division step and conquer step. In the division step, we learn the embedding of training images based on auto-encoder, and partition the training images into adaptive subsets through k-means clustering based on the embedding. In the conquer step, sub-models are fed with subsets separately and trained synchronously. The loss function values of all sub-models are integrated to compute a new weight-sum loss function. The new loss function with tight deviation bound provides more precise gradients for improving performance. Experimental results show that with a fixed number of iterations, DC-MMD-GANs can converge faster, and achieve better performance compared with the standard MMD-GANs on celebA and CIFAR-10 datasets. Full article

Non-intrusive load monitoring (NILM) is a cost-effective approach that electrical appliances are identified from aggregated whole-field electrical signals, according to their extracted electrical characteristics, with no need to intrusively deploy smart power meters (power plugs) installed for individual monitored electrical appliances in a practical field of interest. This work addresses NILM by a parallel Genetic Algorithm (GA)-embodied Artificial Neural Network (ANN) for Demand-Side Management (DSM) in a smart home. An ANN’s performance in terms of classification accuracy depends on its training algorithm. Additionally, training an ANN/deep NN learning from massive training samples is extremely computationally intensive. Therefore, in this work, a parallel GA has been conducted and used to integrate meta-heuristics (evolutionary computing) with an ANN (neurocomputing) considering its evolution in a parallel execution relating to load disaggregation in a Home Energy Management System (HEMS) deployed in a real residential field. The parallel GA that involves iterations to excessively cost its execution time for evolving an ANN learning model from massive training samples to NILM in the HEMS and works in a divide-and-conquer manner that can exploit massively parallel computing for evolving an ANN and, thus, reduce execution time drastically. This work confirms the feasibility and effectiveness of the parallel GA-embodied ANN applied to NILM in the HEMS for DSM. Full article

The autonomous landing of an Unmanned Aerial Vehicle (UAV) on a marker is one of the most challenging problems in robotics. Many solutions have been proposed, with the best results achieved via customized geometric features and external sensors. This paper discusses for the first time the use of deep reinforcement learning as an end-to-end learning paradigm to find a policy for UAVs autonomous landing. Our method is based on a divide-and-conquer paradigm that splits a task into sequential sub-tasks, each one assigned to a Deep Q-Network (DQN), hence the name Sequential Deep Q-Network (SDQN). Each DQN in an SDQN is activated by an internal trigger, and it represents a component of a high-level control policy, which can navigate the UAV towards the marker. Different technical solutions have been implemented, for example combining vanilla and double DQNs, and the introduction of a partitioned buffer replay to address the problem of sample efficiency. One of the main contributions of this work consists in showing how an SDQN trained in a simulator via domain randomization, can effectively generalize to real-world scenarios of increasing complexity. The performance of SDQNs is comparable with a state-of-the-art algorithm and human pilots while being quantitatively better in noisy conditions. Full article

Convolutional neural network (CNN) is well-known for its powerful capability on image classification. In hyperspectral images (HSIs), fixed-size spatial window is generally used as the input of CNN for pixel-wise classification. However, single fixed-size spatial architecture hinders the excellent performance of CNN due to the neglect of various land-cover distributions in HSIs. Moreover, insufficient samples in HSIs may cause the overfitting problem. To address these problems, a novel divide-and-conquer dual-architecture CNN (DDCNN) method is proposed for HSI classification. In DDCNN, a novel regional division strategy based on local and non-local decisions is devised to distinguish homogeneous and heterogeneous regions. Then, for homogeneous regions, a multi-scale CNN architecture with larger spatial window inputs is constructed to learn joint spectral-spatial features. For heterogeneous regions, a fine-grained CNN architecture with smaller spatial window inputs is constructed to learn hierarchical spectral features. Moreover, to alleviate the problem of insufficient training samples, unlabeled samples with high confidences are pre-labeled under adaptively spatial constraint. Experimental results on HSIs demonstrate that the proposed method provides encouraging classification performance, especially region uniformity and edge preservation with limited training samples. Full article

Fast and accurate airport detection in remote sensing images is important for many military and civilian applications. However, traditional airport detection methods have low detection rates, high false alarm rates and slow speeds. Due to the power convolutional neural networks in object-detection systems, an end-to-end airport detection method based on convolutional neural networks is proposed in this study. First, based on the common low-level visual features of natural images and airport remote sensing images, region-based convolutional neural networks are chosen to conduct transfer learning for airport images using a limited amount of data. Second, to further improve the detection rate and reduce the false alarm rate, the concepts of “divide and conquer” and “integral loss’’ are introduced to establish cascade region proposal networks and multi-threshold detection networks, respectively. Third, hard example mining is used to improve the object discrimination ability and the training efficiency of the network during sample training. Additionally, a cross-optimization strategy is employed to achieve convolution layer sharing between the cascade region proposal networks and the subsequent multi-threshold detection networks, and this approach significantly decreases the detection time. The results show that the method established in this study can accurately detect various types of airports in complex backgrounds with a higher detection rate, lower false alarm rate, and shorter detection time than existing airport detection methods. Full article