Indoor sensing technologies are gaining increasing attention in the development of smart environments. Wi-Fi-based human activity recognition, which exploits channel state information (CSI), enables accurate detection of human movements by analyzing signal fluctuations caused by activity, even in complex indoor settings. This passive, device-free approach leverages the ubiquity of Wi-Fi signals, eliminating the need for wearable devices and improving user convenience. However, current Wi-Fi-based sensing systems face several limitations, including suboptimal real-time performance, inefficient resource utilization, and the absence of dedicated hardware platforms. To address these challenges, this article presents a wireless sensing device based on printed circuit board technology, integrated with a CSI-driven system for recognizing indoor human behavior. Experimental results across diverse scenarios demonstrate high recognition accuracy, underscoring the proposed system's potential to improve the efficiency and practicality of wireless sensing technologies in smart environments.

Microseismic monitoring has emerged as a critical technique for exploiting tight reservoirs, particularly those involving hydraulic fracturing, such as shale gas and coalbed methane. The conventional migration-based stacking location method for surface microseismic events relies heavily on the accuracy of the velocity model. However, obtaining an accurate three-dimensional (3D) velocity model is often challenging, prompting the common use of one-dimensional (1D) layered velocity models derived from well-logging data or constant velocity models calibrated through perforation shots. To enhance the precision of microseismic event localization and improve practical applicability, we introduce a refined migration-based stacking location method incorporating two depth-dependent effective parameters: stacking velocity and heterogeneity factor. Two effective parameters were found to describe wave raypath through heterogeneity media, which can be estimated by semblance-based scanning technology. Furthermore, to address potential errors in the velocity model and residual statics arising from topographical variations, we incorporate microseismic event moveout-corrected gathers for residual static corrections. This additional step further refines the accuracy of microseismic event locations. Another advantage of our proposed method is its ability to directly compute the theoretical travel time during the migration-based location process, eliminating the need for precomputing and storing a traveltime table. The efficacy and practicality of our method are demonstrated through applications to both synthetic model data and field data examples.

Accurate source localization and source mechanism estimation in microseismic monitoring critically depend on precise subsurface velocity model reconstruction. Recent advances in data-driven approaches, such as neural network models, have demonstrated promising performance in seismic velocity inversion. However, purely data-driven methods are significantly constrained by their reliance on the completeness of the training dataset. Additionally, the performance of current neural network architectures remains suboptimal and requires further improvement. We propose a step-by-step training methodology that integrates the wave equation-based modeling with a modified U-shaped neural network (U-Net). This approach incorporates physical laws of wave propagation to constrain the updating process of the velocity model. The proposed method carries out the purely data-driven procedure first and then incorporates waveform modeling with finite-difference simulations. Besides, the new method defines dynamic weights within the loss function and introduces the Akaike information criterion (AIC) to detect and extract the P and S phases, which reduces the computational expenses as well as alleviates the artifacts in velocity models effectively. We conducted experiments with synthetic fault models and realistic layered models, and the results show that the wave equation-constrained method recovers the velocity model better than the purely data-driven algorithm. The approach leveraging physics-based and data-driven models can be easily adapted to other geophysical inversion tasks.

Migrated dip-angle gathers provide a visible domain for estimating Fresnel zones of migration. However, it remains a challenge to extract Fresnel zones from the dip-angle gathers with higher precisions and less workloads, especially in 3D cases. A commonly used approach is to yield a pair of 2D dip-angle gathers by stacking the 3D dip-angle gathers along the inline and crossline direction, then to extract the two 1D Fresnel zones from the resulting pair of 2D dip-angle gathers, and finally incorporate the two 1D Fresnel zones into 2D one. The deep learning-based data-driven method has proven efficient in automatically extracting 1D Fresnel zones from a 2D dip-angle gather. However, unlike the current data-driven methods that treat the extraction of each 1D Fresnel zone as an independent task, we propose a unified multi-task learning network based on the segmentation model to simultaneously extract a pair of 1D Fresnel zones from the pair of 2D dip-angle gathers. The network architecture consists of a shared encoder and multi-head decoder modules, thus leveraging shared knowledge and representations across tasks while extracting task-specific features through end-to-end training. The proposed architecture significantly reduces the total number of training parameters and accelerates the training process compared to independent networks for each task (i.e., single-task learning). Furthermore, the network’s generalizability is enhanced compared to single-task learning due to the fact that multiple tasks act as regularizations for each other. We demonstrate the method using a real 3D field data. Higher signal-to-noise ratio (SNR) migration results are obtained.

Time-lapse seismic imaging is an essential and effective tool in characterizing reservoir changes due to oil and gas production or CO2 injection. We herein focus on imaging with the time-lapse difference between observed datasets using reverse-time migration (RTM). RTM has no dip limitation but remains an adjoint imaging operator with Hessian effects, such as amplitude imbalance and blurring effects. Its corresponding inverse operator, least-squares RTM (LSRTM), promises much higher imaging quality but at an expensive cost. To balance computational overhead and imaging quality, we take the one-step LSRTM, which relieves the Hessian effects through a data-domain adaptive deconvolution. We verify the proposed approach on a synthetic dataset from a modified Marmousi model containing three fluid- or gas-related anomalies. The results show that our approach can detect and describe the detailed time-lapse reservoir changes in high resolution.

Seismic migration produces the migrated images of subsurface media using seismic data, which is important for geophysical exploration. However, the adjoint-based migration methods may produce a blurry image, convolved by a Hessian matrix. To address this problem, we propose a Hessian-assisted iterative self-training learning (HAISTL) method aimed at approximating the inverse Hessian matrix and deblurring the migrated image. First, we train a long short-term (LSTM) network using labeled images and use it as a teacher network to generate pseudolabels for the unlabeled images. Subsequently, we integrate the demigration and migration operators to identify the pseudolabels with high confidence levels and construct a dataset containing both the true and pseudolabels. The dataset is then used to train a student network with the injection of model noise into the network. Finally, we regard the student network as a new teacher and repeat the process in an iterative STL framework. We demonstrate the effectiveness of our proposed method using two synthetic datasets and field data. Compared with the supervised learning (SL) method, the proposed method exhibits superior generalization capabilities. This advantage stems from the incorporation of the demigration and migration operators, providing a valuable prior for the inverse Hessian matrix in training the model. In contrast to the model-driven least-squares migration (LSM) methods, the proposed method yields high-resolution images with significantly reduced computational costs. However, it may be less effective in recovering small-scale structures when confronted with an extremely limited number of labels.

Implementing Fresnel zones-based stacking is a pursuit in various seismic imaging methods. However, the estimation of Fresnel zones remains a challenge in three-dimensional migration. Migrated dip-angle gathers provide a visible domain for estimating Fresnel zones. An analytical estimation (i.e., a model-driven method) faces limitations in automatically estimating Fresnel zones in migrated dip-angle gathers in real-world situations due to complex reflections, non-uniform coverage, and noises. Human interaction remains necessary for precisely estimating Fresnel zones in migrated dip-angle gathers. Due to the high number of Fresnel zones in 3D cases, interpolation is necessary to fill in the gaps between manually estimated zones. Aiming to reduce the workload of human interaction and mitigate interpolation errors, we propose a semantic segmentation model (i.e., a deep learning-based data-driven method) to estimate Fresnel zones in migrated dip-angle gathers automatically. We transform the estimation of Fresnel zones into a binary classification task of each pixel in dip-angle gathers. Instead of training the network using the entire dip-angle gather images, we train the network using patches to make the network focus on learning the detailed and general features within the patches. Our proposed network, named deep-supervised attention-UNet, is trained using a deep-supervised method along with a hierarchical hybrid loss function to segment the dip-angle gather on different scales. This approach yields superior segmentation results compared to the UNet model, in qualitative and quantitative aspects. We test the efficiency and practicability of our method using a marine field data set. The signal-to-noise ratio (SNR) of migration results obtained using the Fresnel zones estimated by our method is improved significantly.

Generalized few-shot object detection aims to achieve precise detection on both base classes with abundant annotations and novel classes with limited training data. Existing approaches enhance few-shot generalization with the sacrifice of base-class performance, or maintain high precision in base-class detection with limited improvement in novel-class adaptation. In this paper, we point out the reason is insufficient Discriminative feature learning for all of the classes. As such, we propose a new training framework, DiGeo, to learn Geometry-aware features of interclass separation and intra-class compactness. To guide the separation of feature clusters, we derive an offline simplex equiangular tight frame (ETF) classifier whose weights serve as class centers and are maximally and equally separated. To tighten the cluster for each class, we include adaptive class-specific margins into the classification loss and encourage the features close to the class centers. Experimental studies on two few-shot benchmark datasets (VOC, COCO) and one long-tail dataset (LVIS) demonstrate that, with a single model, our method can effectively improve generalization on novel classes without hurting the detection of base classes. Our code can be found here.

Task offloading in edge computing is important for the Industrial Internet of Things (IIoT) to implement computation-intensive applications in real time. However, achieving efficient task offloading in IIoT is very challenging due to the limited computing resources of IIoT devices, the coupling of computing and communication resources, and the unreliability in multihop wireless transmission. In this article, we construct a link model by considering the influence of unreliable links in multihop transmission to reveal the relationship between reliability and transmission delay. Then, a nonconvex optimization problem that minimizes task processing delay is formulated, and task offloading is decided by considering transmission path selection, bandwidth allocation, and computational resource allocation. To solve this problem, an algorithm based on the alternating direction method of multipliers (ADMM) is designed using auxiliary variables and reformulation linearization technology (RLT). The simulation results show that our proposed algorithm can fully utilize the computing power of the edge server and reduce the task processing delay. Compared with the centralized algorithms, the performance of the proposed scheme is only 1% worse, but the calculation time can be reduced by 40%.

Regression testing is an essential software maintenance activity before the release of a new version implementing a bug fix or a new feature. A regression test selection (RTS) technique chooses a subset of existing test cases to ensure that the system will not be adversely affected by the latest modifications. With the rise of DevOps, behavior-driven development (BDD) is growing in popularity as it is in close alignment with agile practices, for example, continuous integration. Hence, it is necessary to propose a novel and effective RTS technique for BDD specifically to accelerate the development process while ensuring software quality. Since most existing techniques for RTS are code-based and thus subject to some limitations, we present a requirement-based technique which uses the requirements in BDD to select test cases in both high-level (acceptance testing) and low-level (unit testing). Our technique firstly illustrates the new requirement with a scenario, and subsequently computes the semantic similarity of the new scenario and all existing scenarios with the vector space model. According to the results, the modification-traversing regression test cases can be selected in a semi-automated way. We also conduct an experimental study to evaluate our technique in terms of inclusiveness, precision, efficiency and generality. The study shows that our technique is applicable for BDD and effective in practice.

The interpretability of black-box text classification models has been receiving widespread attention in recent years accompanying the growing popularity of artificial intelligence.To garner user trust on the model's decision-making process, it is imperative to provide faithful instance-wise justifications and rationalize the prediction in a human-readable way.In this paper, we address this challenge by introducing Locally Universal Rules (LURs) as model-agnostic local explanations.LURs are a subset of input words sufficient for the model to arrive at a particular prediction, even if the rest of words are perturbed slightly.We show the identification of the optimal LUR is NP-complete.Consequently, we propose a population-based algorithm LUR-Locator to perform the constrained optimization efficiently.We conduct extensive experiments to evaluate our algorithm on a cross product of well-established text classification datasets and models.The empirical results demonstrate that LURLocator can efficiently generate high-quality local explanations, as compared to existing explanatory methods.

To grasp what makes the deep learning models arrive at a particular prediction, gradient-based explanatory methods have been widely used in Natural Language Processing (NLP) recently.While the saliency maps of images can be computed directly in the pixel-level input space, the continuous gradient vector for words has to be reduced to a single value to indicate the word-level importance, and existing methods such as Sensitivity Analysis (SA) and Gradient × Input (GI) are either tricky or short of a deep investigation.In this paper, we review the family of gradient-based explanatory methods and discuss their practical implications.Specially, we propose the signed version of GI, namely SignedGI, while some previous work may have misunderstandings on its signedness.We also show the weakness of SA-based methods.We conduct extensive experiments to evaluate these explanatory methods both qualitatively and quantitatively.

Software defect prediction (SDP) utilizes a trained prediction model to predict the defect proneness of code modules in a software system by mining the inherent characteristics of historical defect data. An effective model can optimize the allocation of testing resources, thus improving the quality of software products. Most previous studies use handcrafted features to represent code snippets, but the main problem is that it is difficult to capture the semantic and structural information of the code context, which is often crucial for software defect prediction. Meanwhile, most of the existing software defect prediction models cannot make predictions at the code line level, which makes it extremely arduous to provide developers with more detailed reference information. To address these issues, in this paper, we propose a model based on ensemble learning techniques and attention mechanisms to offer more comprehensive prediction information to developers by locating suspect lines of code when making method-level defect predictions. This model leverages abstract syntax trees (ASTs) as the intermediate representation of code snippets. Since the historical defect data has a striking characteristic of class-imbalance, an approach based on Self-organizing Map (SOM) clustering is employed to handle noisy data. Experimental results show that, on average, the proposed model improves the F-measure by 17.7% and AUC by 37.8%, compared with the other four machine learning algorithms.

Recently, the development of Industrial Internet of Things has made the emergence of innovative applications which are usually computation-intensive and latency-critical. In this paper, we study the task offloading in software-defined access network, where the industrial devices and edge computing servers are connected to wireless access points. To meet applications' requirements about latency and computing capacity and realize computational load balancing, we formulate a mixed integer non-linear program problem to minimize the overall cost. First, we propose an algorithm based on convex optimization and matching theory to minimize the overhead. Then, we consider load balancing of edge servers and extend our problem with regard to overhead and load balancing cost. We adopt alternating direction method of multipliers algorithm to solve the extended problem. Finally, simulation results demonstrate the effectiveness and superiority of our proposed algorithm.

Due to the nested nonlinear structure inside neural networks, most existing deep learning models are treated as black boxes, and they are highly vulnerable to adversarial attacks. On the one hand, adversarial examples shed light on the decision-making process of these opaque models to interrogate the interpretability. On the other hand, interpretability can be used as a powerful tool to assist in the generation of adversarial examples by affording transparency on the relative contribution of each input feature to the final prediction. Recently, a post-hoc explanatory method, layer-wise relevance propagation (LRP), shows significant value in instance-wise explanations. In this paper, we attempt to optimize the recently proposed explanation-based attack algorithms (EAAs) on text classification models with LRP. We empirically show that LRP provides good explanations and benefits existing EAAs notably. Apart from that, we propose a LRP-based simple but effective EAA, LRPTricker. LRPTricker uses LRP to identify important words and subsequently performs typo-based perturbations on these words to generate the adversarial texts. The extensive experiments show that LRPTricker is able to reduce the performance of text classification models significantly with infinitesimal perturbations as well as lead to high scalability.