This Letter presents a flexible wire-driven linear actuator using a bio-inspired sealing method that allows to replace a rigid rod by a flexible one and reduce the size of a robot link containing pneumatic or hydraulic linear actuators. Conventional pneumatic or hydraulic linear actuator sealing hinders the use of flexible wires as actuator rods. In contrast, the proposed tube sealing resembles a long balloon wrapping around the wire and stretching or contracting with the wire rod motion. Therefore, the wire is separated from the chamber and prevents fluid leakage through the flexible wire rod. This sealing method is inspired by earthworm which has slippery skin and stretching or contracting motion. The authors fabricated a prototype wire rod pneumatic linear actuator using the proposed tube seal and conducted leakage tests. In addition, the authors evaluated linearity between the input pressure and output force to determine the linear actuator performance.

In this Letter, low-pass finite impulse response filters are proposed to flexibly reduce the envelope bandwidth (BW) requirements for envelope tracking (ET) supply modulators. A second-order power ET (PET) approach is used as a comparative study, and a ‘design zone’ for shaping function coefficients is identified for the first time. Proposed 20 and 10 MHz low-pass filters (LPFs) are used to reduce the effective envelope BW in response to a 20 MHz long-term evolution signal applied to a 10 W power amplifier (PA). Experimental results verify that in comparison with PET, the proposed LPF method has the benefit of allowing arbitrary BW reduction and is compatible with the use of customised shaping functions. It is also shown that significant memory effects can be introduced when the effective envelope BW is reduced to less than the RF BW. To investigate the linearisability of the ET PA, the linearity before and after digital pre-distortion is also measured.

Noise is an inherent characteristic of electronic circuits. In this Letter, the authors propose a novel estimation technique for a noise influenced electronic circuit. Stochasticity is accounted for by utilising Itô's stochastic calculus. This novel estimation technique unifies the Carleman linearisation with the Fokker–Planck equation for the Itô stochastic differential equations. To show the effectiveness of the proposed method, numerical simulation results are presented for a benchmark tunnel diode oscillator circuit. This method is beneficial to other electronic circuits for estimating the state trajectories under the influence of randomness when observations are not available.

In this Letter, a reconfigurable processing element (PE) for pipelined SDF FFT architecture is presented, which can be configured to compute 2, 3 and 5-point DFTs. Foremost, the proposed PE architecture for the 5-point DFT computation is designed by factorising the 5-point DFT computation operation into cyclic convolution units and then the 2- and 3-point DFTs structures are mapped on to it using multiplexers. Thus, all three configurations are possible. In the case of prior 5-point PE designs, the PE can start its operation only after the arrival of all the five-input data, whereas the proposed PE completes a part of computation after the arrival of the first three inputs and reuse the same hardware to process the next two inputs. As a result, the proposed PE requires less hardware, at the same time, preserving the throughput of prior PE. The proposed PE required 25% less multiplier and one adder less compared to the Winograd algorithm based 5-input PE.

The authors present a unique application of analogue oxide-based resistive memory (OxRAM) device for sensor-level information storage and computation. They show that quality of low-contrast images in low-light can be improved by carefully exploiting OxRAM conductance modulation from specific bi-layer OxRAM material stacks. The proposed methodology involves conversion of light intensity to pulse frequency followed by resistance encoding as different non-volatile OxRAM resistance states.

Foreground segmentation is to segment moving objects from a video or camera. Many approaches have been proposed to perform background subtraction. However, these algorithms could not handle complex scenes like illumination changes, dynamic background and camouflage etc. In this Letter, the authors propose a robust end-to-end convolutional neural network with multiscale features to address these problems. They use only a few training samples with RGB channels and output the corresponding foreground mask, which could save millions in labelling ground-truth for the big dataset. They evaluate their model using Change Detection 2014 dataset, and results outperformed other state-of-the-art methods with F-measure of 0.982 with ∼30 fps processing speed.

While convolutional neural networks (CNNs) have become the first choice for the medical image segmentation, they still lack the critical ingredient of incorporating priors, such as smoothness and boundary shapes. The authors tackle the limitation by developing a novel prior that is boundary-aware in two ways: promoting smoothness without blurring object boundaries and punishing prediction errors according to boundary shapes. They bring the boundary-aware property into effect by weighting the prediction gradients and errors with the distance map. Their prior differs from previous approaches that either over-smooth boundaries or tend to produce rough boundaries. They evaluate their prior alongside the cross-entropy (CE) on a cardiac MRI dataset. Compared to CE alone, their prior improves the Dice score by 1.5% and Hausdorff distance by 53%. It also yielded a faster and more stable learning process.

Redundant information generated in the process of feature fusion and residual information always remain challenging in multi-person pose estimation. In this Letter, the authors present an optimised cascaded pyramid attention network composed of two novel modules to reduce the redundant information and highlight the residual information for more accurate results. The first module, called channel optimisation module (COM), optimises the channel at different levels. The proposed COM consisting of convolution layers and two pooling layers performs on the feature maps Conv-25 of different resolution to reduce the redundant information after convolution operation. The second module, called residual attention bottleneck module (RABM), highlights the residual information in the bottleneck. Through adopting the attention mechanism in the bottleneck, RABM performs a significant role in reducing the redundant information. In contrast to the previous work, they attach more attention to residual information, helping to tackle the challenging problem for the ‘hard’ keypoints. By merging COM and RABM into Cascaded Pyramid Network, the authors' method achieves competitive results of 74.6 AP on the COCO keypoint benchmark, outperforming the state-of-the-art results.

The source code suggestions provided by current integrated development environment (IDEs) are mostly dependent on static type learning. These suggestions often end up proposing irrelevant suggestions for a peculiar context. Recently, deep learning-based approaches have shown great potential in the modelling of source code for various software engineering tasks. However, these techniques lack adequate generalisation and resistance to acclimate the use of such models in a real-world software development environment. This Letter presents DeepVS, an end-to-end deep neural code completion tool that learns from existing codebases by exploiting the bidirectional gated recurrent unit (BiGRU) neural net. The proposed tool is capable of providing source code suggestions instantly in an IDE by using pre-trained BiGRU neural net. The evaluation of this work is two-fold, quantitative and qualitative. Through extensive evaluation on ten real-world open-source software systems, the proposed method shows significant performance enhancement and its practicality. Moreover, the results also suggest that DeepVS tool is capable of suggesting zero-day (unseen) code tokens by learning coding patterns from real-world software systems.

An accurate exponential histogram test (EHT) method is proposed for the ADC linearity test. Different from conventional EHT methods which estimate ADC static parameters only based on the number of hits, the proposed method accurately estimates the ADC transition levels, differential non-linearity and integral non-linearity by creative utilisation of the information of stimulus signal. As a result, the proposed method avoids the errors introduced by the assumed proportional relationship between the code width and the number of hits in conventional methods. Simulation results show that the proposed method outperforms the conventional methods in terms of accuracy and robustness.

A wideband differential filtering power divider using three-line coupled structure with multiple lumped elements is proposed. The equal power division ratio, filtering response, isolation and common-mode suppression are mainly derived from the symmetrical structure, three-line coupled lines with capacitors, resistors and the short lines along the vertical symmetrical line, respectively. Compared to the reported differential filtering power dividers, the proposed design shows advantages of wide bandwidth, compact size and low insertion loss (IL). A prototype with the small size of 0.5λ _g × 0.08λ _g (λ _g is the guided wavelength at the centre frequency) was designed for demonstration, which exhibits the 3 dB fractional bandwidth of 79.2% and the minimum IL of 0.48 dB.

Powder electroluminescent devices (PELDs) have a high potential for commercial application because of flat light emission devices and simple printing process fabrication with low-cost production, without semiconductor processes using vacuum and plasma. In conventional PELDs, indium tin oxide (ITO) has been used as a transparent electrode. However, ITO is unfavourable in flexible electronic devices due to its intrinsic brittleness. The authors evaluated the current and luminance characteristics of PELDs prepared on gravure offset-printed invisible silver-grid (Ag-grid) electrodes that were laminated with and without poly(3,4-ethylenedioxythiophene):poly(styrenesulphonate) (PEDOT:PSS) transparent electrodes. The PELD on the Ag-grid laminated with PEDOT:PSS electrode shows the maximum luminance of ∼250 cd/m² at 200 μm spacing, which is five-fold greater than the luminance of PELD without PEDOT:PSS electrode lamination.

One of the serious limitations in frequency-modulated continuous wave light detection and ranging (FMCW-LiDAR) is the maximum ranging distance as imposed by the coherence of light sources. The author proposes and demonstrates a new technique for overcoming the limitation using differential detection scheme between the frequency-fixed carrier and frequency-modulated subcarrier, which are generated from a single laser source. In the proof-of-principle experiment, two reflection points with a 3 m separation are clearly resolved at distances 192.2 m and 1.504 km using a distributed feedback (DFB) laser with the coherence length of 70 m.

In this Letter, a battery state of health monitoring technique suitable for battery management systems is presented. This method analyses switching and battery transitions in the battery management system topologies to distinguish any early degraded battery cell in the battery strings. The proposed technique processes the voltage of the system, captured by an already employed sensor during the transition of the states. To verify the performance of the proposed technique, some experiments have been carried out on a typical charge balancing circuit.

Sparse representation-based classification (SRC) possesses remarkable characteristics for application in synthetic aperture radar (SAR) automatic target recognition (ATR), for instance, inherent feature extraction and robustness to articulation, and so on. However, the performance of SRC is highly sensitive to parameters such as sufficient training samples, SAR images quality, and targets’ changing conditions in depression, pose, configuration, and so on. Unfortunately, the training sample resources for SAR ATR are often expensive and scarce. Further, the targets in SAR images, even with slightest changes in conditions, display mutable characteristics attributable to unique SAR image formation, and speckle noise corruption. To overcome these obstacles, this Letter proposes to establish several compact and complementary dictionaries using monogenic signal's components of SAR images and Fisher discriminative dictionary learning. Then, an optimal decision fusion (ODF) strategy is proposed, which utilises SRC and the latter dictionaries for robust SAR ATR. Compared with single classifiers or multiple parallel classifiers, the proposed ODF increases the accuracy of recognition, while at the same time, decreases the complexity of the system. The proposed methods have considerable features making them applicable in practical situations. Based on the experimental results, the proposed methods outperform state-of-the-art approaches.

Highly sampled imagery offers many benefits to the radar practitioner, ranging from easier image coregistration to simple visual appeal. However, it is often overlooked due to the computational burden forming such an image imposes. Fast image formation typically imposes restrictions on the imaging scenario, e.g. synthetic aperture radar (SAR) far-field, and exploits parallelism through the use of modern multi-core architecture. Imposing a SAR near-field requirement on the image formation limits the applicability of several faster algorithms, thus there is a need to create a general process to achieve highly sampled imagery, regardless of the imaging regime. In this Letter, a method for accurately upsampling near-field (SAR) imagery is presented. This is applicable to both SAR near-field and SAR far-field scenarios. The methodology is discussed, and an example is provided in the form of a SAR near-field volumetric image of a miniature tank. The limitations to the approach are discussed and prospects for future work given.

Normalised subband adaptive filtering algorithms have attracted attention due to their ability to present faster convergence in the case of coloured excitation data. The NSAF-SF scheme is an example of a state-of-the-art subband adaptive algorithm that demands a low computational burden. This Letter proposes a deterministic local optimisation approach with affine constraints whose result leads to an enhanced NSAF-SF updating mechanism. One of these constraints consists of a projection into a hyperplane derived from a relaxed -norm restriction, which provides a sparsity-promoting scheme. Such a constraint incorporates prior information into the (originally sparsity-agnostic) NSAF-SF. Such prior information concerns the energy concentration of the ideal transfer function that the adaptive filter intends to identify. Furthermore, the advanced optimisation problem is modified in order to make the advanced adaptive filter robust against impulsive noise. The simulations present a performance improvement for both transient and steady-state regions.

Today, music is a very important and perhaps inseparable part of people's daily life. There are many genres of music and these genres are different from each other, resulting in people to have different preferences of music. As a result, it is an important and up-to-date issue to classify music and to recommend people new music in music listening applications and platforms. Classifying music by their genre is one of the most useful techniques used to solve this problem. There are a number of approaches for music classification and recommendation. One approach is based on the acoustic characteristics of music. In this study, a music genre classification system and music recommendation engine, which focuses on extracting representative features that have been obtained by a novel deep neural network model, have been proposed. Acoustic features extracted from these networks have been utilised for music genre classification and music recommendation on a data set.

In this Letter, the authors investigate a modified encoding scheme of batched sparse (BATS) codes to improve the erasure floor performance. In the proposed regularised variable-node (RBATS) codes, variable-node degrees are regularised to maximise the minimum degree for transmission. The asymptotical decoding performance of RBATS codes is analysed by a tree-based analysis. Furthermore, they optimise check-node degree distribution of RBATS codes to decrease decoding overhead according to the decoding condition. It is demonstrated through simulation results that the optimised RBATS codes can not only lower the transmission overhead but also improve erasure floor performance compared with traditional BATS codes.

This Letter proposes a symbol cyclic reconstruction scheme for universal-filtered multi-carrier (UFMC) systems. In this scheme, the overlap-add technique is employed to reduce the computational complexity of the system transmitter. At the receiver, ISI cancellation is performed to remove the interference between neighbouring overlapped symbols. Then, a cyclic reconstruction algorithm is proposed to eliminate the interference caused by the linear shift. This scheme effectively improves the transmission efficiency in multipath channel environments while ensuring the bit error rate performance. Simulation results show that the TE of the proposed scheme is much higher than that of standard UFMC and cyclic prefix -OFDM systems.