Search

Currently, weakly supervised semantic segmentation methods based on image-level annotation often rely on pseudo pixel masks generated by seed regions. However, the growth of seed regions is stochastic, and in cases where targets are occluded or overlapped in the image without additional reference information, the segmentation network may encounter issues of missed or incorrect segmentation. To address this problem, this paper proposes a polarized attention mechanism for weakly supervised semantic segmentation networks. The attention mechanism consists of a semantic perception branch and a boundary detection branch. The semantic perception branch allows the network to better distinguish the category of each pixel in the image. Subsequently, the boundary detection branch enables the seed region to naturally expand towards the target boundary. The pseudo pixel mask generated by this method provides better coverage of the target area and improves the performance of the segmentation network. The test set and validation set mean Intersection over Union (mIoU) of the PASCAL VOC 2012 dataset achieved 72.1% and 73.2%. The results of the experiments demonstrated the effectiveness of the proposed method. The experimental results indicated that the attention mechanism, as proposed in this paper, can effectively enhance the segmentation performance in situations where objects in the image are occluded or overlapped.

An electromagnetic (EM) logging while drilling (LWD) tool is proposed for deep reading detection ahead of the drill bit. The EM LWD tool consists of one coil transmitter sub and two separated coil receiver subs. We utilize the attenuations of measured voltage signals as indicators for detecting the boundary ahead of the drill bit. The effects of frequency and the tool length on the detection capability of the tool are discussed. Also, the attenuations of coaxial voltage signals and coplanar voltage signals proactively respond to the anisotropic formation boundary ahead of the drill bit. The attenuations of coaxial and coplanar signals are sensitive to variations of horizontal conductivity of a conductive anisotropic target layer, but not sensitive to variations in its vertical component. Moreover, an air hang test verifies that the depth of detection of the tool could reach more than 30 m, which is also confirmed by the numerical simulations with an air-seawater-sediment model. The proposed EM LWD tool for detecting the formation boundary ahead of the drill bit is promising for applications of deep reading detections.

Named Entity Recognition (NER) is a fundamental but crucial task in natural language understanding, aiming at identifying entity mentions from free text. Current methods mainly use sequence-labeling and span-based models, where the former ignores the importance of token interaction, and the latter pays little attention to the global inter-dependency among entity tokens. In this work, we propose a novel NER model that consists of two branches: a Token-Pair Interaction Module (TPIM) and a U-shaped Network. The TPIM models head-tail relations between token pairs while capturing intrinsic token connectivity within entity boundaries. The U-shaped Network is employed to capture the contextual dependency in the token-pair relation matrix. Furthermore, we build a typical domain-specific entity dataset CCAEE based on real-world applications in the chemical accident domain. The experimental results on CCAEE and CLUENER datasets demonstrate the effectiveness of our proposed model.

In order to solve the problems of semantic loss and inaccurate boundary detection in the process of object tracking, a visual tracking algorithm combining parallel structure with dual attention-aware mechanism is proposed in this paper. As backbone network, parallel structure is composed of Convolutional neural network and Attention Cooperative(CAC) processing module, which is used for feature extraction. Because this structure can capture the local and global information of the target at the same time, it can solve the problem of semantic information loss. Dual Attention-aware Network(DAN) is used for feature enhancement, which is composed of target-aware attention and boundary-aware attention. Template online updating strategy is used to improve template quality, and an effective score prediction module-Template Elimination Mechanism(TEM) is designed in the CAC processing module to select high quality templates. This kind of object tracking algorithm which combines local and global information is called TrackCAC. The evaluation results on different datasets show that the algorithm can maintain high tracking precision and success in different scenarios. It shows good robustness and accuracy in the performance evaluation results on VOT datasets.

This paper presents a Lightweight Dense Convolutional (LDC) neural network for edge detection. The proposed model is an adaptation of two state-of-the-art approaches, but it requires less than 4% of parameters in comparison with these approaches. The proposed architecture generates thin edge maps and reaches the highest score (i.e., ODS) when compared with lightweight models (models with less than 1 million parameters), and reaches a similar performance when compare with heavy architectures (models with about 35 million parameters). Both quantitative and qualitative results and comparisons with state-of-the-art models, using different edge detection datasets, are provided. The proposed LDC does not use pre-trained weights and requires straightforward hyper-parameter settings. The source code is released at https://github.com/xavysp/LDC.

With the new age of data innovation, the Internet of Things (IoT) proliferation has drawn enormous thought and has applied to help applications in different fields i.e., natural assurance, military observing, and industrial applications. WSNs are the essential segment of IoT for monitoring as well as tracking. The most preeminent applications provide confinement and identification of continuous objects i.e. wildfire, toxic gas, bio synthetics concoctions, and so forth. In the case of continuous objects such as fire and toxic gases are detected to identify the boundary of damage and alert teams for rescue efforts. It is also helpful in identifying safe paths for rescue. We have investigated various existing surveys that carried out different concepts associated with continuous object tracking and find out the deficit of boundary detection of object. In order to replete the present cleft of analysis, we have inspected various current state-of-the-art works on boundary detection of a continuous object that has yet not been added to the current writing. This paper presents an extensive overview of different continuous object tracking schemes which involve energy efficiency, boundary detection, communication, data aggregation, and network structural design in literature with the aid of featuring taxonomy. We summarized, compared, and classified these schemes along with their analysis and performance. Moreover, for further evaluation mechanism, strengths and weaknesses of these schemes are presented. Finally, various state-of-the-art open research challenges are identified. Moreover, there is a need to overcome these challenges through novel and reliable arrangements by the researchers.

Along with a large and increasing number of resource-constrained and battery-powered smart things being deployed in networks, an energy-efficient mechanism for continuous object boundary detection is fundamental for prolonging the network lifetime. Most of the existing studies focus on how to improve the boundary accuracy of continuous objects while ignoring the energy consumption of smart things in edge networks, which may lead to some smart things being in an unusable state. To address this issue, we propose an ${E}$ nergy-efficient ${C}$ ontinuous object boundary ${D}$ etection ${M}$ echanism, namely ECDM. This mechanism consists of two stages: (i) edge-collaboration monitoring mechanism, where only sensory data relevant to toxic gas are transmitted among multi-edge networks, thus minimizing the number of exchanged sensory data involved in the detection, and reducing the transmission energy consumption of smart things; (ii) gas diffusion model is used to predict the diffusion of toxic gases, such that suspicious edge networks can be located quickly and efficiently, further reducing the energy consumption of smart things transmitting anomalous sensory data. Experimental results show that the energy consumption of ECDM is reduced by 34.5% compared to the latest boundary detection method that introduces a gas diffusion model while maintaining detection accuracy.

Internet of Things (IoT) has been widely used to facilitate environmental perception, where detecting the boundary regions for continuous objects with energy-efficient manner is a challenge to be explored to support domain applications. This paper proposes a novel approach for continuous objects boundary prediction and detection in IoT sensing network, which is called Cloud Model-IoT Sensing Network Collaborative (CM-IoTSNC). Specifically, when an event is potentially occurred, the atmospheric dynamic diffusion model deployed on the cloud is adopted to predict gas diffusion trend in ideal and complex environments, moreover prediction results are transmitted to the IoT devices in the real-time fashion for scheduling security plans in advance. One-hop neighbor nodes are activated by abnormal nodes to determine a more accurate boundary region. Compared our technique with two traditional methods, namely Wireless Sensor Monitoring and Activating One-hop Neighbor Nodes, experimental results show that our method has a good performance in reducing the energy consumption and prolonging the lifetime of the network.

The shape, size and distribution of grains play an important role in the quality of alloy materials. However, in the actual production process, the grain boundary is not always visible in the metallographic image observed by microscope. In this paper, we propose an end-to-end deep network named GF-RCF based on adversary network and feature learning to detect grain boundaries effectively. The network is guided by a novel multi-level loss through its three parts: base network, adversary network and metric module, which are used for boundary detection, boundary inpainting and feature learning respectively. Specifically, for base network, its architecture is based on richer convolution feature to achieve basic boundary detection by pixel-level loss. For adversary network, it employs the image-level adversary loss and feature matching loss to repair the invisible boundary. For metric module, it involves a feature learning strategy by comparing the features of unlabeled data with the labeled data to improve the generalization of the network. The multi-level loss refers to the total loss of pixel-level, image-level and feature-level obtained from the three subnetwork, respectively. From the perspective of machine learning, GF-RCF involves semi-supervised learning, transfer learning and generative adversarial network, which makes our model have more powerful learning ability. In experiments, GF-RCF achieves 0.8792 F1 score, surpassing other methods and reaching the state of the art.

By using local and global image information, a novel active contour method based on the p-Laplace equation for image segmentation is proposed in this paper. The force term in the evolution equation incorporates global, local and edge information of the image. The global and local terms drag the contour towards object boundaries precisely and take care of the contour movement when it is away from the object. Meanwhile, the integration of edge stopping function in this force term ensures the stoppage of contour at object boundaries to avoid boundary leakage problem. The variable exponent p-Laplace energy is used for smoothness of the level set to detect the exact object boundaries in the presence of complex topological changes and deep depression. Finally, the adaptive force and p-Laplace energy term are jointly integrated into a level set by using a simple finite difference scheme to build the final evolution equation for the method. The proposed method has strong capability to accurately segment the images having noise, intensity inhomogeneity, and complex object boundaries. Moreover, the proposed method overcomes the problem arise from contour initialization as the evolution of contour is independent of the initialization of level set function. Experimental results on synthetic, real and medical images along with two publicly available databases validate the robustness and effectiveness of the proposed method.

Clustering algorithms have a very wide range of applications on data analysis, such as machine learning, data mining. However, data sets often have problems with unbalanced and non-spherical distribution. Clustering by fast search and find of density peaks (DPC) is a density-based clustering algorithm which could identify clusters with non-spherical data. In real applications, this algorithm and its variants are not very effective for the division of unevenly distributed clusters, because they only use one indicator (the distance of neighbor points) to handle inner points and boundary points at the same time. To this end, we introduce a new indicator named asymmetry measure which enhances the ability of finding boundary points. Then we propose a boundary detection-based density peaks clustering (BDDPC) algorthm that combines the above two indicators, so that different clusters are separated from each other accurately and the purpose of improving the clustering effect is achieved. The BDDPC algorithm can not only cluster uniformly distributed data, but also cluster unevenly distributed data. In real life, the distribution of high-dimensional data sets are always unbalanced, so this algorithm has very important applications. Experimental results with synthetic and real-world data sets illustrate the effectiveness of our algorithm.

This paper describes a mixed-signal electrocardiogram (ECG) system for personalized and remote cardiac health monitoring. The novelty of this paper is fourfold. First, a low power analog front end with an efficient automatic gain control mechanism, maintaining the input of the ADC to a level rendering optimum SNR and the enhanced recyclic folded cascode opamp used as an integrator for $\Sigma \Delta $ ADC. Second, a novel on-the-fly PQRST boundary detection (BD) methodology is formulated for finding the boundaries in continuous ECG signal. Third, a novel low-complexity ECG feature extraction architecture is designed by reusing the same module present in the proposed BD methodology. Fourth, the system is having the capability to reconfigure the proposed low power ADC for low (8 b) and high (12 b) resolution with the use of the feedback signal obtained from the digital block when it is in processing. The proposed system has been tested and validated on patient’s data from PTBDB, CSEDB, and in-house IIT Hyderabad Data Base (IITHDB) and we have achieved an accuracy of 99% upon testing on various normal and abnormal ECG signals. The whole system is implemented in 180-nm technology resulting in 9.47- $\mu \text{W}$ (at 1 MHz) power consumption and occupying 1.74- $mm^{2}$ silicon area.