Search

Semi-supervised learning has been proven to be effective in utilizing unlabeled samples to mitigate the problem of limited labeled data. Traditional semi-supervised learning methods generate pseudo-labels for unlabeled samples and train the classifier using both labeled and pseudo-labeled samples. However, in data-scarce scenarios, reliance on labeled samples for initial classifier generation can degrade performance. Methods based on consistency regularization have shown promising results by encouraging consistent outputs for different semantic variations of the same sample obtained through diverse augmentation techniques. However, existing methods typically utilize only weak and strong augmentation variants, limiting information extraction. Therefore, a multi-augmentation contrastive semi-supervised learning method (MAC-SSL) is proposed. MAC-SSL introduces moderate augmentation, combining outputs from moderately and weakly augmented unlabeled images to generate pseudo-labels. Cross-entropy loss ensures consistency between strongly augmented image outputs and pseudo-labels. Furthermore, the MixUP is adopted to blend outputs from labeled and unlabeled images, enhancing consistency between re-augmented outputs and new pseudo-labels. The proposed method achieves a state-of-the-art performance (accuracy) through extensive experiments conducted on multiple datasets with varying numbers of labeled samples. Ablation studies further investigate each component's significance. [ABSTRACT FROM AUTHOR]

To address the problem that the performance of the detector in airborne magnetic anomaly detection (MAD) is terrible, a stochastic resonance (SR) detection algorithm based on orthonormalized basis function (OBF-SR) is proposed for MAD under low signal-to-noise ratio conditions. The signal contaminated by noise is first preprocessed by the OBF method, where the sum of the three components in the OBF space is selected as the SR system input. Then, a parallel SR system with different initial states is designed to detect the signal. Finally, the simulation analysis of MAD methods is performed to draw a comparison between the OBF-SR method, the typical SR method, and the OBF method. The results show that the OBF-SR method outperforms the SR and OBF methods in the detection probability and detection range under the same conditions. [ABSTRACT FROM AUTHOR]

In practical operation, the system parameters of a permanent magnet linear motor are affected by unknown factors, including nonlinear friction, sudden load changes, thrust fluctuations, and so on. Therefore, a controller designed for fixed parameters often cannot produce satisfactory results. To solve this problem, a novel design method for permanent magnet linear motor systems based on backstepping is proposed in this article. The proposed control scheme does not require the values or range of changes of the system parameters in advance but constructs an update law to perform online parameter estimation. A difficulty encountered in controller design and stability analysis is the estimation of the unknown coefficient of a single-variable state, which can be considered as a virtual input. In this article, the estimated virtual control coefficient is introduced into the coordinate transformation to construct a novel coordinate transformation. However, introducing the estimated coefficient into the virtual control makes the derivative of the Lyapunov function used in the stability analysis more complex. This novel nonlinear dynamic term can be precisely cancelled by changing the update law of the coefficient. Finally, simulations are performed to assess the performance of the proposed control scheme against that of a traditional proportional–integral–derivative controller, and the simulation results show that the proposed control law can effectively improve the system performance. [ABSTRACT FROM AUTHOR]

This paper addresses the issue of the reference signal being contaminated by target echoes due to factors such as wide beamwidths, beam pointing errors, or site errors in passive bistatic radar (PBR). Such contamination can lead to reference signal impurity and subsequently degrade the performance of traditional object detection methods. To overcome this challenge, a novel target detection method centered around the reference signal's purification is presented. The proposed method involves the modeling, analysis, localization, and purification of the impure reference signal. First, a new signal model incorporating the impure reference signal is established, followed by a comprehensive analysis of its impact on target detection. Then, leveraging the valuable information hidden in the impure reference signal, target echoes within the impure reference signal are localized and reconstructed. Finally, the reconstructed target echoes are subtracted from the original impure reference signal to purify the reference signal. Based on the purified reference signal, traditional target detection is performed to detect the expected targets. The core advantage of this method lies in its efficient reconstruction and removal of the impure terms within the reference signal. The effectiveness of the proposed method is demonstrated through a series of simulation results. [ABSTRACT FROM AUTHOR]

The modeling of channel and temporal information is of crucial importance for action recognition tasks. To build a high-performance action recognition network by effectively capturing channel and temporal information, we propose CLS-Net: an action recognition algorithm based on channel-temporal information modeling. The proposed CLS-Net characterizes channel and temporal information by inserting multiple modules to an end-to-end backbone network, including a channel attention module (CA module) for modeling channel information, a long-term temporal module (LT module) and a short-term temporal module (ST module) for modeling temporal information. Specifically, the CA module extracts the correlation between feature channels so the network can learn to selectively strengthen the features containing useful information and suppress the useless features through global information. The LT module moves some channels in the temporal dimension to realize information interaction across time domains and model global temporal information. The ST module enhances the motion-sensitive features by calculating the feature-level frame difference information and realizes the representation of local motion information. Since the multi-module insertion mode directly affects the whole model's final performance, we propose a novel multi-module insertion mode instead of a simple series or parallel connection to ensure that the multiple modules can complement one another and cooperate with each other more efficiently. CLS-Net achieves SOTA performance on the EgoGesture and Jester dataset in the same type of network and achieves competitive results on the Something-Something V2 dataset. [ABSTRACT FROM AUTHOR]

Sodium metal batteries (NMBs) have attracted increasing attention as next‐generation rechargeable batteries. How to improve their cycling stability and safety under limited sodium excess conditions, ideally zero sodium excess (i.e., anode‐free architecture), is highly desired yet remains challenging. Herein, it is demonstrated that sodium formate (HCOONa), one component of the solid electrolyte interphase (SEI) naturally formed on sodium metal anode, is a promising candidate for designing high‐performance artificial SEI layers, which can suppress the sodium dendrite formation and reduce the side reactions between sodium and the electrolyte. Profiting from the HCOONa interface, the Na|Na3V2(PO4)3 battery with a high mass loading of Na3V2(PO4)3 (10 mg cm−2) exhibits a superior cycling stability with an ultralow decay rate of 0.004% per cycle over 800 cycles. More impressively, a single molecular layer of HCOONa in situ formed on commercial copper current collector helps to extend the lifespan of the anode‐free Cu|Na3V2(PO4)3 battery to 400 cycles with 88.2% capacity relation, representing the longest cycle lifetime reported in anode‐free NMBs. [ABSTRACT FROM AUTHOR]

In order to improve the efficiency of human–robot interaction (HRI), it is necessary to carry out research on precise control of the manipulator. In this paper, an adaptive non-singular fast terminal sliding mode control scheme is proposed for robot manipulators to solve the trajectory tracking problem with model uncertainty and external disturbances. At first, a novel non-singular fast terminal sliding mode surface is proposed, and by introducing an auxiliary function, the singularity problem caused by the inverse of the error-related matrix could be avoided in the controller design process. Then, the controller is developed by using Lyapunov synthesis. A robust adaptive strategy is used to deal with lumped uncertainty, with an adaptive update law designed to compensate for the upper bound of lumped uncertainty whose upper bound is prior unknown. Finally, a two-link robot manipulators as a simulation example is given to illustrate the effectiveness of the proposed scheme. Compared with other similar algorithms, the proposed adaptive non-singular fast terminal sliding mode control scheme has higher efficiency and smaller computational complexity for the reason that no piecewise continuous function is needed to be constructed during the controller design. [ABSTRACT FROM AUTHOR]

The segmentation of pulmonary lobes is important in clinical assessment, lesion location, and surgical planning. Automatic lobe segmentation is challenging, mainly due to the incomplete fissures or the morphological variation resulting from lung disease. In this work, we propose a learning-based approach that incorporates information from the local fissures, the whole lung, and priori pulmonary anatomy knowledge to separate the lobes robustly and accurately. The prior pulmonary atlas is registered to the test CT images with the aid of the detected fissures. The result of the lobe segmentation is obtained by mapping the deformation function on the lobes-annotated atlas. The proposed method is evaluated in a custom dataset with COPD. Twenty-four CT scans randomly selected from the custom dataset were segmented manually and are available to the public. The experiments showed that the average dice coefficients were 0.95, 0.90, 0.97, 0.97, and 0.97, respectively, for the right upper, right middle, right lower, left upper, and left lower lobes. Moreover, the comparison of the performance with a former learning-based segmentation approach suggests that the presented method could achieve comparable segmentation accuracy and behave more robustly in cases with morphological specificity. [ABSTRACT FROM AUTHOR]

Organic room‐temperature long‐persistent luminescent materials are promising light‐emitting materials for encryption, architectural decoration, organic solar cells, and biomedical applications. However, their unstable structures and thermal‐ and humidity‐induced emission quenching have greatly limited their utility and reliability. Here, we report a metal‐free nonconjugated copolymer that possesses stable photoluminescence at both high temperature and humidity. The room‐temperature long‐persistent luminescence (LPL) of this copolymer lasts for more than 15 s and can be recovered in high humidity conditions by heating to remove moisture. Copolymer LPL can be achieved with various excitation wavelengths, ranging from ultraviolet to near‐infrared, and the LPL color can be adjusted accordingly. The high initial LPL intensity and ultrafast filling time of the copolymer makes it suitable for low flicker alternating current‐driven light‐emitting diodes (AC‐LEDs). [ABSTRACT FROM AUTHOR]

This article deals with the problem of electromagnetic source localization (ESL). An evolutionary particle filter, which is first used to make a decision on the positions of electromagnetic sources, has two characteristics. One characteristic is that the number of particles can be significantly reduced while the other characteristic is that the particle diversity can be well improved. On the basis of the estimated positions of electromagnetic sources, the position and velocity of the virtual leader can be determined. Then, an event-based fixed-time consensus control approach is proposed such that the positions and velocities of robots reach consensus with the virtual leader over a fixed-time interval while saving resource consumption by reducing the communication frequencies and updating times of control inputs. Finally, simulation and experimental results show the effectiveness of the proposed decision and event-based fixed-time consensus control approach for ESL. [ABSTRACT FROM AUTHOR]

All‐solid‐state thin‐film lithium batteries (TFLBs) are the ideal wireless power sources for on‐chip micro/nanodevices due to the significant advantages of safety, portability, and integration. As the bottleneck for increasing the energy density of TFLBs, the key components of cathode, electrolyte, and anode are still underway to be improved. In this review, a brief history of TFLBs is first outlined by presenting several TFLB configurations. Based on the state‐of‐the‐art materials developed for lithium‐ion batteries (LIBs), the challenges and related strategies for the application of those potential electrode and electrolyte materials in TFLBs are discussed. Given the advanced manufacture and characterization techniques, the recent advances of TFLBs are reviewed for pursuing the high‐energy‐density and long‐term‐durability demands, which could guide the development of future TFLBs and analogous all‐solid‐state lithium batteries. [ABSTRACT FROM AUTHOR]

One open issue of target detection for synthetic aperture radar (SAR) images is the capture effect from the clutter edge and the interfering outliers, including surrounding targets in the multitarget environment, sidelobes, and ghosts. To address this issue, a superpixel-level constant false-alarm rate (CFAR) detector is proposed based on the truncated Gamma statistics for the multilook intensity SAR data. Superpixel segmentation serves as a preprocessing procedure to divide the SAR image into meaningful patches. By automatic clutter truncation in the superpixel-level background clutter window, the real clutter samples are preserved. The experimental results with real SAR images demonstrate that the proposed method achieves better goodness-of-fit performance for the real clutter background with outlier exclusion, yielding a higher target detection rate in the multitarget environments. [ABSTRACT FROM AUTHOR]

Compared with uniform arrays, a generalized sparse array (GSA) can obtain larger array aperture because of its larger element spacing, which improves the accuracy of DOA estimation. At present, most DOA estimation algorithms are only suitable for the uniform arrays, while a few DOA estimate algorithms that can be applied to the GSA are unsatisfactory in terms of computational speed and accuracy. To compensate this deficiency, an improved DOA estimation algorithm which can be applied to the GSA is proposed in this paper. First, the received signal model of the GSA is established. Then, a fast DOA estimation method is derived by combining the weighted noise subspace algorithm (WNSF) with the concept of "transform domain" (TD). Theoretical analysis and simulation results show that compared with the traditional multiple signal classification (MUSIC) algorithm and the traditional WNSF algorithm, the proposed algorithm has higher accuracy and lower computational complexity. [ABSTRACT FROM AUTHOR]

Solid‐state lithium batteries (SSLBs) are promising owing to enhanced safety and high energy density but plagued by the relatively low ionic conductivity of solid‐state electrolytes and large electrolyte–electrode interfacial resistance. Herein, we design a poly(vinylidene fluoride‐co‐hexafluoropropylene) (PVDF‐HFP)‐based polymer‐in‐salt solid electrolyte (PISSE) with high room‐temperature ionic conductivity (1.24×10−4 S cm−1) and construct a model integrated TiO2/Li SSLB with 3D fully infiltration of solid electrolyte. With forming aggregated ion clusters, unique ionic channels are generated in the PISSE, providing much faster Li+ transport than common polymer electrolytes. The integrated device achieves maximized interfacial contact and electrochemical and mechanical stability, with performance close to liquid electrolyte. A pouch cell made of 2 SSLB units in series shows high voltage plateau (3.7 V) and volumetric energy density comparable to many commercial thin‐film batteries. [ABSTRACT FROM AUTHOR]

Polarized smart antenna array has attracted considerable interest due to its capacity of matched reception or interference suppression for active sensing systems. Existing literature does not take full advantage of the combination of polarization isolation and smart antennas and only focuses on uniform linear array (ULA). In this paper, an innovative synthesis two-dimensional beampattern method with a null that has cross-polarization for polarized planar arrays is proposed in the first stage. This method aims to further enhance the capability of interference suppression whose optimization problem can be solved by second-order conic programming. In the second stage, a new sparse array-optimized method for the polarized antenna array is proposed to reduce the high cost caused by the planar array that is composed of polarized dipole antennas. Numerical examples are provided to demonstrate the advantages of the proposed approach over state-of-the-art methods. [ABSTRACT FROM AUTHOR]

Abstract: To address the serious capacity fading in lithium–sulfur batteries, a 3D ferroconcrete‐like aminated carbon nanotubes network with polyaniline coating as an effective sulfur host to contain polysulfide dissolution is presented here. In this composite, the cross‐linked aminated carbon nanotubes framework provides a fast charge transport pathway and enhancement in the reaction kinetics of the active material to greatly improve the rate capability and sulfur utilization. The ethylenediamine moieties provide strong adhesion of polar discharge products to nonpolar carbon surfaces and thus efficiently prevent polysulfide dissolution to improve the cycle stability, confirmed by density functional theory calculations. The outside polyaniline layers structurally restrain polysulfides to prevent the shuttle effect and active material loss. Benefiting from these advantages, the synthesized composite exhibits a high initial capacity of 1215 mAh g−1 and a capacity of 975 mAh g−1 after 200 cycles at 0.2 C. Even after 200 cycles at 0.5 C, a capacity of 735 mAh g−1 can be maintained, among the best performance reported. The strategy in this work can shed some light on modifying nonpolar carbon surfaces via the amination process to chemically attach sulfur species for high‐performance lithium–sulfur batteries. [ABSTRACT FROM AUTHOR]

On-chip integrated solenoid inductors with nanogranular FeCoTiO magnetic cores have been fabricated. Both single-layered and multilayered magnetic cores with different geometries were studied. An inductance of 14.2 nH was achieved when using a multilayered core structure and keeping the device area of 0.14 mm2. The inductance of the magnetic core inductor is $\sim 14$ times that of the air core inductor, and the quality factor is 7.5. The permeability trends calculated from inductance were compared with the permeability spectra of the patterned FeCoTiO films measured by an improved shorted microstrip transmission line perturbation method. Consistent results have been obtained after considering the demagnetization effect of the magnetic core with high permeability. [ABSTRACT FROM PUBLISHER]

In this article, a new morphing method is proposed for two linear features at different scales, based on their entire structures ( MLBES in abbreviation). First, the bend structures of the linear features are identified by using a constrained Delaunay triangulation ( CDT in abbreviation) model and represented by binary bend-structure trees. By matching the independent bends represented by the bend-structure trees, corresponding independent bends are obtained. These corresponding independent bends are further used to match their child bends based on hierarchical bend structures so that corresponding bends are obtained. On this basis, the two linear features are split into pairs of corresponding subpolylines by the start and end points of the corresponding bends. Second, structures of the corresponding subpolylines are identified by the Douglas- Peucker algorithm and represented by binary line generalization trees ( BLG-trees in abbreviation). The corresponding subpolylines are split into smaller corresponding subpolylines by matching the nodes of the BLG-trees. Third, the corresponding points are identified by using the linear interpolation algorithm for every pair of corresponding subpolylines. Finally, straight-line trajectories are employed to generate a family of intermediate-scale linear features. By comparison with other methods, it is found that MLBES is accurate and efficient. [ABSTRACT FROM AUTHOR]