Search

Existing artificial neural network-based methodologies for salient object detection in RGB-depth (RGB-D) images typically require significant memory and computation time. In this paper, we propose ReBiT-Net, an novel and resource-efficient network designed to addresses this issue. ReBiT-Net utilizes a mobile network for feature extraction and incorporates depth map quality to regulate the fusion of multi-modal features, resulting in top-to-bottom refinement of salient objects using salient information. Empirical evaluations conducted on five benchmark datasets demonstrate the superior performance of our model in terms of accuracy (achieving 334 frames per second for an input size of 320 ×320) and model parameters (merely 5.1 MB). Moreover, we introduce ReBiT-Net*, a simplified variant of ReBiT-Net, which entails reduced model parameters (4.2 MB) and enhanced processing speed (793 frames per second for a 256 ×256 input size). These improvements are achieved through reduced memory requirements and computational demands via the adoption of a smaller input image size.

Structure determines the properties. However, whether electronic structure determines geometry or geometry determines electronic structure seems a philosophical question in a chicken and egg situation, which remains unclear. In this work, by applying density functional theory (DFT) and DMRG(4n,4n)-CASSCF methods, theoretical investigation suggested that the dual antiaromaticity in cyclo[2n]carbons with even nshould be attributed to the electron correlation effect, instead of decreased geometric symmetry, which actually exists in all cyclo[2n]carbon molecules and does not point out the essence. Such dual antiaromaticity can be conceptualized as electron correlation-stabilized dual antiaromaticity. Results also showed that DFT is reliable for cyclocarbons larger than C14, but we should be careful when applying it to smaller ones. DFT failed to give the correct structure of C6compared with density matrix renormalization group results.

Non-invasive, closed-loop brain modulation offers an accessible and cost-effective means of evaluating and modulating one’s mental and physical well-being, such as Parkinson’s disease, epilepsy, and sleep disorders. However, wearable EEG systems pose significant challenges for the analog front-end (AFE) circuits in view of µV-level EEG signals of interest, multiple sources of interference, and ill-defined skin contact. This paper presents a direct-digitization AFE tailored for dry-electrode scalp EEG recording, characterized by wide input dynamic range (DR) and high input impedance. The AFE utilizes a second-order 5-bit delta-delta sigma (Δ-ΔΣ) ADC to shape DC electrode offset (DEO) and low-frequency disturbances while retaining high accuracy. A non-inverting pseudo-differential instrumentation amplifier (IA) embedded in the ADC ensures high input impedance (Zin) and common-mode rejection ratio (CMRR). Fabricated in a standard 0.18-μm CMOS process, the AFE delivers 700-mVpp input signal range, 95.3-dB DR, 87-dB SNDR, and 800-MΩ input impedance at 50 Hz while consuming 88.4µW from a 1.2 V supply. The benefits of high DR and high input impedance have been validated by dry-electrode EEG measurement.

Enhancing proton storage in the zinc-ion battery cathode material of MnO2holds promise in promoting its electrochemical performance by mitigating the intense Coulombic interaction between divalent zinc ions and the host structure. However, challenges persist in addressing the structural instability caused by Jahn–Teller effects and accurately modulating H+intercalation in MnO2. Herein, the doping of high-electronegativity Sb with fully occupied d-orbital in MnO2is reported. The Sb doping strategy engenders the formation of Mn–O–Sb path in the structure with a strong dipole polarization field, which facilitates the delocalization of egorbital electron in Mn and thus mitigates the Jahn–Teller effects. Simultaneously, adjusting the level of Sb doping in MnO2leads to modulation of the p-band center of O, optimizing its interaction with hydrogen and thereby enhancing proton storage. Consequently, MnO2doped with 6% Sb exhibits commendable performance in both rate capability and cycling endurance, delivering 113 mAh g–1at 2 A g–1after 2000 cycles. This investigation underscores the crucial role of electronic structural engineering in elevating the electrochemical performance of cathode materials for zinc-ion batteries.

Cancer development takes 10 to 50 years, and epigenetics plays an important role. Recent evidence suggests that ~80% of human cancers are linked to environmental factors impinging upon genetics/epigenetics. Because advanced metastasized cancers are resistant to radiation/chemotherapeutic drugs, cancer prevention by relatively nontoxic “epigenetic modifiers” will be logical. Many dietary phytochemicals possess powerful antioxidant and anti-inflammatory properties that are hallmarks of cancer prevention. Dietary phytochemicals can regulate gene expression of the cellular genome viaepigenetic mechanisms. In this review, we will summarize preclinical studies that demonstrate epigenetic mechanisms of dietary phytochemicals in skin, colorectal, and prostate cancer prevention. Key examples of the importance of epigenetic regulation in carcinogenesis include hypermethylation of the NRF2 promoter region in cancer cells, resulting in inhibition of NRF2-ARE signaling. Many dietary phytochemicals demethylate NRF2 promoter region and restore NRF2 signaling. Phytochemicals can also inhibit inflammatory responses viahypermethylation of inflammation-relevant genes to block gene expression. Altogether, dietary phytochemicals are excellent candidates for cancer prevention due to their low toxicity, potent antioxidant and anti-inflammatory properties, and powerful epigenetic effects in reversing procarcinogenic events.

The Python-based program, XMECP, is developed for realizing robust, efficient, and state-of-the-art minimum energy crossing point (MECP) optimization in multiscale complex systems. This article introduces the basic capabilities of the XMECP program by theoretically investigating the MECP mechanism of several example systems including (1) the photosensitization mechanism of benzophenone, (2) photoinduced proton-coupled electron transfer in the cytosine–guanine base pair in DNA, (3) the spin-flip process in oxygen activation catalyzed by an iron-containing 2-oxoglutarate-dependent oxygenase (Fe/2OGX), and (4) the photochemical pathway of flavoprotein adjusted by the intensity of an external electric field. MECPs related to multistate reaction and multistate reactivity in large-scale complex biochemical systems can be well-treated by workflows suggested by the XMECP program. The branching plane updating the MECP optimization algorithm is strongly recommended as it provides derivative coupling vector (DCV) with explicit calculation and can equivalently evaluate contributions from non-QM residues to DCV, which can be nonadiabatic coupling or spin–orbit coupling in different cases. In the discussed QM/MM examples, we also found that the influence on the QM region by DCV can occur through noncovalent interactions and decay with distance. In the example of DNA base pairs, the nonadiabatic coupling occurs across the π–π stacking structure formed in the double-helix system. In contrast to general intuition, in the example of Fe/2OGX, the central ferrous and oxygen part contribute little to the spin–orbit coupling; however, a nearby arginine residue, which is treated by molecular mechanics in the QM/MM method, contributes significantly via two hydrogen bonds formed with α-ketoglutarate (α-KG). This indicates that the arginine residue plays a significant role in oxygen activation, driving the initial triplet state toward the productive quintet state, which is more than the previous knowledge that the arginine residue can bind α-KG at the reaction site by hydrogen bonds.

Although different damage-coupled physically based models have been proposed, those damage models are established based on the isothermal uniaxial or multi-axial tensile condition. However, the influence of the stress state on ductile failure strain is not adequately considered in the thermal damage models. It is widely recognized that the stress state has significant effect on damage evolution and ductility fracture for metallic materials. Actually, the current damage-coupled physically based models are not suitable for complex hot forming and the applications of these damage-coupled models are mainly focused on sheet hot stamping forming to date. Therefore, a modified damage-coupled unified model considering a broad range of stress states was established to figure out the complex damage behavior and microstructure evolution of metallic materials during the thermal deformation process. In this study, the Mg-6Gd-5Y-0.3Zr alloy was used as the experimental material. First, the flow behavior of the alloy was explored using a series of experiments under various stress states (i.e., uniaxial tensile, shear and compression) under various combinations of strain rates and temperatures. On this basis, a modified damage-coupled physics-based model was developed, in which the dynamic recrystallization, damage, stress triaxiality and Lode parameter were included as internal state variables. Furthermore, the model was integrated into ABAQUS to verify the applicability using hot spinnability test because the stress state was very complex during the forming process. The comparison between the calculated results and experiments is carried out. According to the simulation result, the predicted ultimate thinning rate is 71.14%, with an error of only 7.85% from the experimentally measured ultimate thinning rate of 77.2%, and the correlation coefficient and error for thickness and profile between the predicted values and experimental values are 0.9980 and 5.05%, respectively. These demonstrate that the developed model is reliable.

Shaft seal covers a very wide range of machinery and applications. Lip seals, which are used for sealing shafts against the exit or ingress of liquids, are a type of shaft seal. Reducing the wear of the sealing ring and extending its service life can increase the time interval for machine maintenance and inspection. In this paper, a disk structure for reducing the pressure difference between the inside and outside of the lip seals is proposed. The disk structure forms a layer of rotating oil film on the seal ring, so the oil pressure near the seal ring is less than in a sealing chamber. First, based on the similarity principle and the hypothesis of von Kármán, a theoretical model of the rotating oil film was established, and the dominant influence of inertia of circumferential velocity and viscosity of radial velocity on pressure distribution was expounded. Then, an experimental device that can measure pressure distribution and temperature was constructed, and a pressure distribution measurement experiment was conducted. Based on theoretical models and experiments, it was proved that the rotating disk could generate a pressure drop and the pressure drop was a function of the square of the rotating speed. The influence of temperature and oil film thickness was studied, and it was found that the increase in the temperature and oil film thickness weakened rotational flow inertia and enhanced radial flow viscosity, increasing pressure within the oil film. Additionally, it was observed that the oil in the sealing chamber produced a sunken liquid level due to the stirring of the rotating disk, which would reduce the boundary pressure at the periphery of the oil film and produce a greater pressure drop inside the oil film.

This article describes an ultra-high input-impedance active electrode (AE) circuit and the system to sense biopotential signals through a capacitively coupled interface. Various techniques from both circuit and system aspects are used to eliminate the input parasitic capacitance of the AE. On-chip parasitic capacitance is compensated by an auto-calibrated positive feedback loop (PFL) without applying any reference signal. A capacitor down-scaling technique combined with SAR-assisted PFL calibration enables femtofarad-level resolution of the capacitor array, alleviating the practical constraints of the conventional PFL to implement a small capacitor below 10 fF. Besides, a dummy input structure ensures that the pad and electrostatic discharge (ESD) capacitances are also canceled by the PFL, while off-chip parasitic capacitance on the printed circuit board (PCB) is nulled by active shielding. Fabricated in a standard 0.18- $\mu \text{m}$ 1P6M CMOS process, the AE achieves an ultra-high input impedance of 136 $\text{G}\Omega $ at 60 Hz (average of 10 samples). This is equivalent to an input capacitance of 19.5 fF and corresponds to a $2.7\times $ – $68\times $ improvement over the state-of-the-art. The AE exhibits an input signal range of 700 mVpp and an input-referred noise of $0.72 \mu \text{V}_{\mathrm {rms}}$ (0.5–100 Hz) while consuming $10.46 \mu \text{W}$ from a 1.2-V supply. Each AE integrates an IC and a 3-cm2 copper electrode on the PCB, and the wearable system prototypes successfully measured high-quality electrocardiogram (ECG) and electroencephalogram (EEG) signals from test subjects through capacitively coupled interfaces.

This article presents a fully integrated digital polar transmitter (DPTX) incorporating an efficiency-enhanced digital power amplifier (DPA) and a high-resolution digital-to-time converter (DTC). The proposed single-ended Doherty (SED) PA topology with switched-capacitor (SC) arrays can generate four power efficiency peaks that significantly improve the power-added efficiency (PAE) at the power amplifier (PA)’s power back-offs (PBOs). In addition, this work utilizes a three-segment DTC, built with a delay-locked loop (DLL), to improve the phase resolution and enable real-time process-voltage-temperature (PVT) calibration of the DTC. Fabricated in a 28 nm CMOS technology, the DPTX achieves a peak output power ( $P_{\mathrm {out}}$ ) of 27.7 dBm and a peak PAE of 33.4% at a 2.1 GHz carrier frequency. The PAEs at 2.5, 6, and 12 dB PBOs are 31.1%, 24.5%, and 18%, respectively. With the 2.1 GHz carrier, the DTC achieves 360° full output phase range, ±1.3° differential nonlinearity (DNL), ±1.7° integral nonlinearity (INL), and −133.86 dBc/Hz phase noise at a frequency offset of 1 MHz, while the power dissipation is only 14 mW. When testing the 10 MHz (and 20 MHz) long term evolution (LTE) 64 QAM signals with a 6.2 (5.0) dB peak-to-average power ratio (PAPR), the DPTX achieves an average $P_{\mathrm {out}}$ of 21.5 (22.7) dBm, an average PAE of 23.5% (24.6%), an average system efficiency (SE) of 22.5% (23.8%),–30.6 (–27.7) dB error vector magnitude (EVM), and–34.2/–32.6 (–30.1/–29.2) dBc adjacent channel leakage ratio (ACLR). Note that these results have been achieved with only the ON-chip digital pre-distortion (DPD) algorithm.

This article describes a power-efficient, high dynamic range (DR) incremental ADC (IADC) for wearable biopotential signals recording, where DC and low-frequency disturbances such as electrode offset, 50/60 Hz interference and motion artifact must be tolerated. To achieve a wide DR, the IADC performs a three-step conversion by combining zoom-SAR and extended counting (EC) on top of a second-order incremental delta-sigma modulator (ΔΣM). The hybrid architecture notably reduces the oversampling ratio (OSR) with respect to conventional incremental ΔΣMs, while using the EC further improves the Signal-to-Noise-and-Distortion Ratio (SNDR) by 7.4 to 25.6 dB. Fabricated in a 0.18-μm CMOS technology, the IADC achieves 107.6-dB DR, 104.9-dB peak SNR, and 99.3-dB peak SNDR at 2 kS/s while dissipating 130 μW from 1.8-V (analog) / 1.2-V (digital) supply. This translates to a highly competitive FoMDR of 176.5 dB. The high-DR IADC reduces the gain of the preceding instrumentation amplifier (IA) such that significant DC and low-frequency disturbances can be tolerated. The advantages of high DR have been demonstrated by wearable Electrocardiography (ECG) and Electroencephalography (EEG) recordings under motion artifact.

Evaluating the computational accuracy of Spiking Neural Network (SNN) implemented as in-situ learning on large-scale memristor crossbars remains a challenge due to the lack of a versatile model for the variations in non-ideal memristors. This brief proposes a novel behavioral variation model along with a four-stage pipeline for physical memristors. The proposed variation model combines both absolute and relative variations. Therefore, it can better characterize different memristor cycle-to-cycle (C2C) variations in practice. The proposed variation model has been used to simulate the behavior of two physical memristors. Adopting the non-ideal memristor model, the trace-based spiking-timing dependent plasticity (STDP) unsupervised in-memristor learning system is simulated. Although the synaptic-level weight simulation shows a performance degradation of 7.99% and 4.07% increase in the relative root mean square error (RRMSE), the network-level simulation results show no accuracy loss on the MNIST benchmark. Furthermore, the impacts of absolute and relative C2C variations on network performance are simulated and analyzed through two sets of univariate experiments.

Cuticular wax, a critical defense layer for plants, remains a relatively unexplored factor in rumen fermentation. We investigated the impact of cuticular wax on rumen fermentation using triticale as a model. In total, six wax classes were identified, including fatty acids, aldehydes, alkane, primary alcohol, alkyresorcinol, and β-diketone, with low-bloom lines predominated by 46.05% of primary alcohols and high-bloom lines by 35.64% of β-diketone. Low-wax addition (2.5 g/kg DM) increased the gas production by 19.25% (P< 0.05) and total volatile fatty acids by 6.34% (P> 0.05), and enriched key carbohydrate-fermenting rumen microbes like Saccharofermentans, Ruminococcus, and Prevotellaceae, when compared to non-wax groups. Metabolites linked to nucleotide metabolism, purine metabolism, and protein/fat digestion in the rumen showed a positive correlation with low-wax, benefiting rumen microbes. This study highlights the intricate interplay among cuticular wax, rumen microbiota, fermentation, and metabolomics in forage digestion, providing insights into livestock nutrition and forage utilization.

This article offers an integrated solution for first-person shooter (FPS) games to train agents with adaptive strategies. Solving such complex decision tasks requires generalization ability and adaptive strategies. We develop a framework using a novel adaptive strategic control algorithm combined with advanced techniques, such as the hindsight experience replay, multiagent reinforcement learning, and league training. The approach adopts a multistage learning scheme, consisting of learning a goal-conditioned navigation policy, then transferring to learn sophisticated shooting skills by playing against a league of players, and finally learning adaptive strategies. Our agent achieves the SOTA result in past ViZDoom AI Competitions, surpassing previous top-ranked agents (never seen during training) by a large margin. We provide comprehensive analysis and experiments to elaborate the effect of each component in affecting the agent performance and demonstrate that the proposed and adopted techniques are essential to achieve superior performance in ViZDoom Competition and potentially valuable for general end-to-end FPS games.

This brief describes a power-efficient time-domain analog front-end (AFE) for autonomous sensor nodes. This AFE combines an input pulse-width modulator (PWM) and a gated ring oscillator (GRO) to facilitate low-voltage analog building blocks. The input PWM is implemented with pseudo-differential comparators for voltage-to-time conversion, achieving high input impedance and linearity concurrently. A non-equidistant GRO works at a fixed frequency and quantizes the PWM output in time domain, realizing the first-order noise shaping. The AFE is implemented in a $0.18\mu \text{m}$ standard CMOS process. Clocked at 20kHz (500Hz bandwidth), the AFE achieves 81.6dB peak SNDR, $540{\mathrm{ mV}}_{\mathrm{ pp}}$ input signal range, and 80M $\Omega $ input impedance. The power consumption is $2.2\mu \text{W}$ at 0.8V analog and 0.6V digital supply voltages, and these correspond to a Schreier FoM of 165.2dB. Finally, the AFE properties are verified through several types of sensors.

A wearable and remote fetal monitoring system, FetalCare, is specifically developed for home use. It comprises a stretchable on-skin patch and a wearable pod to capture, process, and transmit fetal signals. Data received by a Bluetooth dongle are processed by edge computing to remove diverse noises and interferences and then displayed by a fetal monitoring platform integrated with cloud services. The highly integrated system can simultaneously measure up to 6-channel electrohysterography (EHG) and 3-channel fetal electrocardiogram (FECG) signals from the abdomen using just a 4-channel data acquisition (DAQ) chip. Electrically conductive composites (ECCs) and silicone implement the stretchable on-skin electrode. The compact wearable pod achieves low input noise of $0.191 \mu $ Vrms (250 Hz bandwidth) while consuming 56 mW and can operate continuously for up to 36 h. We use edge computing to process fetal signals, which utilize fast independent component analysis (FastICA) and empirical mode decomposition (EMD) to obtain high-quality FECG and EHG signals. The fetal monitoring platform based on InfluxDB and Grafana supports real-time, remote monitoring with low latency and historical data retrieval. Finally, we used the contraction consistency index (CCI) to assess the EHG and used metrics such as sensitivity (Se), positive predictive accuracy (PPV), accuracy (ACC), and their harmonic mean (F1) to assess the FECG. The CCI for EHG signals measured by the FetalCare system is 0.8, which implies a highly consistent result with the clinical CTG equipment. The average Se, PPV, ACC, and F1 scores for FQRS complexes extraction are 97.97%, 98.62%, 96.58%, and 98.25%, respectively.

The scaled process technology and the increased operational frequency deteriorated the stability of the Static Random-Access Memory (SRAM) working in the harsh radiation environment, partially due to the Single Event Transient (SET) inducing operational failure. As analyzed in the brief, the SET induced “0-1” upset at the word-line (WL) is the crucial operational failure reason. A Feedback-Split-Gate (FSG) voter without delay penalty at the output is proposed to migrate the SET generated in the decoding array. The FSG-voter with less logic effect reduces path effect of the decoder hardened by Double Mode Redundancy (DMR). SET-sensitive area of the decoder is reduced as a result. A new local reading and writing structure with less latency is proposed to protect cells from SET induced disturbance. The correction coverage cost of the proposed peripheral circuit is 746% improved.

This article presents a time-domain-controlled (TDC) single-inductor triple-output (SITO) step-up converter for portable devices with a bipolar power supply. A time-domain controller is utilized in the proposed converter to achieve a minimized delay time and thus ensure a high-conversion ratio (CR) with a switching frequency of 2 MHz. The proposed power converter employs a time-domain energy distribution (ED) balance scheme to ensure the bipolar outputs are symmetrically regulated. In addition, with the assistance of the output offset correction circuits, the load regulation of the TDC-SITO converter is greatly improved. By using an analog-to-digital converter (ADC)-assisted fast-settling loop, the converter achieves an excellent load transient response with low undershoot/overshoot voltages. Powered by a lithium-ion battery, the TDC-SITO converter can generate a bipolar power supply of $\pm$12 V and a standard 5 V output voltage, indicating a maximum CR of 8.8×. Implemented in a 180 nm bipolar-CMOS-DMOS (BCD) process, the converter achieves a peak conversion efficiency of 91.6% at an output power of 2.25 W. With a load current step of 100 mA between two output voltages of $\pm$10 V, the output voltages recover their normal values within 100 $\mu$s while the undershoot/overshoot voltages are below 160 mV. The maximum output power of the TDC-SITO converter is 9.6 W.

Post-training quantization (PTQ) has been proven an efficient model compression technique for Convolution Neural Networks (CNNs), without re-training or access to labeled datasets. However, it remains challenging for a CNN accelerator to fulfill the efficiency potential of PTQ methods. A large number of PTQ techniques blindly pursue high theoretic compression effect and accuracy, ignoring their impact on the actual hardware implementation, which causes more hardware overhead than benefit. This paper introduces ASLog, a PTQ-friendly CNN accelerator that explores four key designs in an algorithm-hardware co-optimizing manner: the first practical 4-bit logarithmic PTQ pipeline SLogII, the multiplier-free arithmetic element (AE) design, the energy-efficient bias correction element (BCE) design, and the per-channel quantization friendly (PCF) architecture and dataflow. The proposed SLogII PTQ pipeline can push the limit of logarithmic PTQ to 4-bit with < 2.5% accuracy degradation on various image classification and face recognition tasks. Exploiting the approximate computing design and a novel encoding and decoding scheme, the proposed SLogII AE is >40% lower in power and area consumption compared with a common 8-bit multiplier. The BCE and PCF design proposed in this paper are the first to consider the hardware impact of the widely-used per-channel quantization and bias correction technique, enabling an efficient PTQ-friendly implementation with a small hardware overhead. The ASLog is validated in a UMC 40-nm process, with 12.2 TOPS/W energy efficiency and 0.80 mm2 core area. The ASLog can achieve 336.3 GOPS/mm2 area efficiency and >500 OPs/Byte operational intensity, which map to over $1.85\times $ and $1.12\times $ improvement compared with the previous related works.

Envelope tracking (ET) and average power tracking (APT) have been widely used as supply modulators of power amplifiers (PAs). To meet the new challenges brought by the 5G NR standard, the symbol power tracking (SPT) technique is proposed, which is more efficient than ET and more flexible than APT. For the application of subcarrier spacing (SCS) of 15-~60kHz in the frequency range 1 (FR1), this paper proposed a buck-boost SPT supply modulator with a transition time below 1.17- $\mu \text{s}$ for 5G NR PA in the smartphone. The combination of three techniques improves the voltage transition speed. The charge-pump-based switching power modulator increases the output voltage higher than the battery voltage and offers a high voltage source for the auxiliary current paths (ACP). The ACP introduces additional currents to charge/discharge the 1- $\mu \text{F}$ output capacitor. The end-point-prediction (EPP) scheme lets the compensator convert directly from one steady state to the next. This supply modulator circuit is designed and fabricated using 0.18- $\mu \text{m}$ BCD technology. The output voltage range is 1–5-V. The maximum transition speed can be up to 0.37- $\mu \text{s}$ /V and 0.15- $\mu \text{s}$ /V for up-tracking and down-tracking, respectively.

The memristor has been extensively used to facilitate the synaptic online learning of brain-inspired spiking neural networks (SNNs). However, the current memristor-based work can not support the widely used yet sophisticated trace-based learning rules, including the trace-based Spike-Timing-Dependent Plasticity (STDP) and the Bayesian Confidence Propagation Neural Network (BCPNN) learning rules. This paper proposes a learning engine to implement trace-based online learning, consisting of memristor-based blocks and analog computing blocks. The memristor is used to mimic the synaptic trace dynamics by exploiting the nonlinear physical property of the device. The analog computing blocks are used for the addition, multiplication, logarithmic and integral operations. By organizing these building blocks, a reconfigurable learning engine is architected and realized to simulate the STDP and BCPNN online learning rules, using memristors and 180 nm analog CMOS technology. The results show that the proposed learning engine can achieve energy consumption of 10.61 pJ and 51.49 pJ per synaptic update for the STDP and BCPNN learning rules, respectively, with a 147.03× and 93.61× reduction compared to the 180 nm ASIC counterparts, and also a 9.39× and 5.63× reduction compared to the 40 nm ASIC counterparts. Compared with the state-of-the-art work of Loihi and eBrainII, the learning engine can reduce the energy per synaptic update by 11.31× and 13.13× for trace-based STDP and BCPNN learning rules, respectively.

Driver errors, such as distraction, perceptual blindness, and incorrect control manipulation, can either cause road accidents or reduce driving performance in daily driving tasks. Several works in literature have illustrated perceptual blindness and distraction are associated with insufficient attention to those activities vital for safe driving. Also, inappropriate driving-related eye movements may subsequently result in manipulation errors, such as inappropriate control of the throttle or steering wheel. Although many studies have examined drivers' visual performance, there have been few attempts to compare left-hand traffic (LHT) and right-hand traffic (RHT) conditions. Even driving in the same driving scene, different eye movement patterns may also be induced due to drivers' left gaze bias. Motivated by this human factor, this research investigates potential human errors in driving from the perspective of eye movements (i.e., saccades and fixations) and corresponding control manipulations. Firstly, a driving simulator is used to investigate the discrepancy of eye movements in identical driving scenes between LHT and RHT. Secondly, visual regions and feature sensitivities are spatially correlated by comparing the gaze distribution from the central to peripheral visual fields for both LHT and RHT. Besides, considering the violation of traffic rules are likely the cause of potentially fatal driving eye movement and control mistakes, a recommendation strategy is proposed to alleviate visual errors by utilizing a neural network inspired by the driver gaze map. The experiments in both LHT and RHT driving tasks demonstrate improvements in the alleviation of human errors by importing the proposed eye movement recommendation strategy.

Continuous and precise code delay tracking is essential in challenging environments such as urban canyons where multipath and frequent signal blockage may lead to performance degradation and loss of lock of signals. The strobe correlator (SC) is an effective technique for baseband multipath mitigation and is commonly used for GNSS signal tracking enhancement. An SC has two pairs of early and late correlators, one with a wide and another with a narrow correlator spacing. This article extends the conventional SC design by applying its correlator pairs on dual-band GNSS signals to realize an extended strobe correlator (ESC). An ESC takes advantage of the natural distinct bandwidth of narrow-band and wide-band signals, e.g., L1 C/A and L5Q signals, to form a dual-band combination at the correlator level in a deeply coupled manner. Narrow-band L1 signal tracking ensures robustness, while the wide-band L5 signal improves the tracking precision. A moving average window filters the L1 measurement to prevent the noise impact on L5 tracking, providing necessary interband assistance to resolve challenges such as code delay misalignment and precision mismatch. Through both Monte Carlo simulation and real data collected in urban environments, this work demonstrates that an ESC can achieve improved overall robustness against loss-of-lock and multipath mitigation performance over standalone tracking. Compared with KF-based combined (KFC-based) tracking, ESC has higher tracking precision under weak signal environment and robustness during signal blockage, but larger residual multipath error.

Recent existing methods generally adopt a simple concatenation or addition strategy to integrate features at the fusion layer, failing to adequately consider the intrinsic characteristics of different modal images and feature interaction of different scales, which may produce a limited fusion performance. Toward this end, we introduce a cross-scale iterative attentional adversarial fusion network, namely CrossFuse. More specifically, in the generator, we design a cross-modal attention integrated module to merge the intrinsic content of different modal images. The parallel spatial-independent and channel-independent pathways are proposed to calculate the attentional weights, which are assigned to measure the activity levels of source images at the same scale. Moreover, we construct a cross-scale iterative decoder framework to interact with different modality features at different scales, which can constantly optimize their activity levels. By this means, the generator learns to integrate their modality characteristics via attentional weights in an iterative manner, and the generated result characterizes competitive infrared radiant intensity and distinct visible detail description. Extensive experiments on three different benchmarks demonstrate that our CrossFuse outperforms other nine state-of-the-art methods in terms of fusion performance, generalization ability and computational efficiency. Our codes will be released at https://github.com/Zhishe-Wang/CrossFuse.

After implantation, complex and highly specialized molecular events render functionally distinct organ formation, whereas how the epigenome shapes organ-specific development remains to be fully elucidated. Here, nano-hmC-Seal, RNA bisulfite sequencing (RNA-BisSeq), and RNA sequencing (RNA-Seq) were performed, and the first multilayer landscapes of DNA 5-hydroxymethylcytosine (5hmC) and RNA 5-methylcytosine (m5C) epigenomes were obtained in the heart, kidney, liver, and lung of the human foetuses at 13–28 weeks with 123 samples in total. We identified 70,091 and 503 organ- and stage-specific differentially hydroxymethylated regions (DhMRs) and m5C-modified mRNAs, respectively. The key transcription factors (TFs), T-box transcription factor 20 (TBX20), paired box 8 (PAX8), krueppel-like factor 1 (KLF1), transcription factor 21 (TCF21), and CCAAT enhancer binding protein beta (CEBPB), specifically contribute to the formation of distinct organs at different stages. Additionally, 5hmC-enriched Alu elements may participate in the regulation of expression of TF-targeted genes. Our integrated studies reveal a putative essential link between DNA modification and RNA methylation, and illustrate the epigenetic maps during human foetal organogenesis, which provide a foundation for for an in-depth understanding of the epigenetic mechanisms underlying early development and birth defects.

Graphical abstract:

This study reports a carbonylative acetylation for the synthesis of N-phenyl-N-(pyridin-2-yl)acetamides using N,N-dimethylformamide (DMF) as a methyl source and CO as a carbonyl source. Interestingly, dimethyl sulfoxide (DMSO) can be also used as a methyl source when using only DMSO as the solvent. Mechanistic studies using DMSO-d6revealed that the methyl group was derived from the methyl group of DMF instead of DMSO when using DMF and DMSO as a mixed solvent. These results indicated that DMF was a preferential methyl source.

In this letter, we investigate the performance of Rate-Splitting Multiple Access (RSMA) for an uplink communication system with finite blocklegnth (FBL). Considering a two-user Single-Input Single-Output (SISO) Multiple Access Channel (MAC), we study the impact of blocklength and target rate on the throughput and error probability performance of RSMA where one user message is split. We demonstrate that RSMA can outperform Non-Orthogonal Multiple Access (NOMA) in terms of throughput and error probability performance.

Enzyme-free DNA strand displacement process is often practical when detecting miRNAs expressed at low levels in living cells. However, the poor kinetics, tedious reaction period, and multicomponent system hamper its in vivo applications to a great extent. Herein, we design a branch-shaped trapping device (BTD)-based spatial confinement reactor and applied it for accelerated miRNA in situ imaging. The reactor consists of a pair of trapped probe-based catalyzed hairpin assembly (T-CHA) reactions attached around the BTD. The trapping device naturally offered CHA reactions a good spatial-confinement effect by integrating the metastable probes (MHPa and MHPb) of the traditional CHA with the four-branched arm of BTD, which greatly improved the localized concentration of probes and shortened their physical distance. The autonomous and progressive walk of miRNA on the four-arm nanoprobes via T-CHA can rapidly tie numerous four-arm nanoprobes into figure-of-eight nanoknots (FENs), yielding strong fluorescence that is proportional to the miRNA expression level. The unique nanoarchitecture of the FEN also benefits the restricted freedom of movement (FOM) in a confined cellular environment, which makes the system ideally suitable for in situ imaging of intracellular miRNAs. In vitro and in situ analyses also demonstrated that the T-CHA overall outperformed the dissociative probe-based CHA (D-CHA) in stability, reaction speed, and amplification sensitivity. The final application of the T-CHA-based four-arm nanoprobe for imagings of both cancer cells and normal cells shows the potential of the platform for accurately and timely revealing miRNA in biological systems.

This article describes a high-input-impedance, low-noise bioimpedance (BioZ) sensor interface IC for small-area dry-electrode cardio-respiratory signals monitoring. To facilitate high-precision BioZ sensing with high-impedance dry electrodes, the IC utilizes three key techniques as follows: 1) a bias control loop (BCL) to eliminate the excitation current mismatch, reducing the voltage fluctuation on high-impedance input; 2) a quiet-chopping current feedback instrumentation amplifier (QC-CFIA) to mitigate the input-signal-dependent noise; and 3) a full pre-charge (FPC) technique to cancel the input parasitic capacitance for impedance boosting. Manufactured in a 0.18- $\mu \text{m}$ CMOS process, the BioZ prototype IC occupies an area of 0.4 mm2 while consuming 15.8- and 14.4–128.4- $\mu \text{W}$ current, respectively, from the amplifier and the excitation current generator (CG). With these proposed techniques, this IC achieves a high input impedance of 100 $\text{M}\Omega $ at 50 kHz, 0.5- $\text{m}\Omega /\surd $ Hz sensitivity at 1 Hz, and a 106-dB signal-to-noise ratio (SNR). A gel-free respiration and impedance cardiography (ICG) recording has been successfully demonstrated on the human body with four 0.45-cm2 dry electrodes.

Envelope tracking is an effective way to improve the efficiency of radio frequency power amplifiers (PAs). This article presents a wideband hybrid envelope tracking supply modulator (ETSM) for 5G new radio PAs. To reach a wide envelope tracking bandwidth of 200 MHz while maintaining high power efficiency, this work utilized a linear amplifier with accurate frequency compensation and enhanced driving capability. The proposed low-pass sense and control circuit ensures that the switching power amplifier only tracks the low-frequency components of the envelope to reduce the switching power loss. The proposed dual-path duty modulator regulates the voltage across the ac-coupling capacitor. This supply modulator circuit is designed and fabricated using 0.18 μm BCD technology. When tracking 200 MHz bandwidth envelope signals, the prototype ETSM achieves 78.5% efficiency at the output power of 2.7 W when taking a realistic 4 Ω//600 pF load as the PA model.

This article presents a chopper-stabilized multipath current-feedback instrumentation amplifier (CFIA) that can also be reconfigured to a general-purpose operational amplifier (OPA). By utilizing a local positive feedback loop (LPFL), both the CFIA and the OPA benefit from gain-independent impedance boosting while retaining the merits of chopping, i.e., low noise and low offset. In addition, the chopping-induced ripple, one of the main issues of conventional chopper amplifiers, is suppressed by a passive high-pass filter with the SAR-assisted offset calibration. The prototype IC is implemented in a 0.18- $\mu \text{m}$ CMOS process and draws 1.07-mA current from a 5-V supply. The chopper CFIA exhibits the minimal input impedance of 1.8 $\text{G}\Omega $ and the maximum input referred ripple of 0.15 $\mu \text{V}$ (ten samples for both tests). These correspond to at least 1.13 $\times $ and 1.3 $\times $ improvements on the state-of-the-art IAs, respectively. The CFIA and the OPA also achieve excellent power efficiency (the highest GBW/ $\text{I}_{\mathrm {SUPPLY}}$ ), competitive input offset (4 and 2 $\mu \text{V}$ ), and input noise density (39 and 20 nV/ $\surd $ Hz).

Driver identification has been popular in the field of driving behavior analysis, which has a broad range of applications in anti-thief, driving style recognition, insurance strategy, and fleet management. However, most studies to date have only researched driver identification without a robust verification stage. This paper addresses driver identification and verification through a deep learning (DL) approach using psychological behavioral data, i.e., vehicle control operation data and eye movement data collected from a driving simulator and an eye tracker, respectively. We design an architecture that analyzes the segmentation windows of three-second data to capture unique driving characteristics and then differentiate drivers on that basis. The proposed model includes a fully convolutional network (FCN) and a squeeze-and-excitation (SE) block. Experimental results were obtained from 24 human participants driving in 12 different scenarios. The proposed driver identification system achieves an accuracy of 99.60% out of 15 drivers. To tackle driver verification, we combine the proposed architecture and a Siamese neural network, and then map all behavioral data into two embedding layers for similarity computation. The identification system achieves significant performance with average precision of 96.91%, recall of 95.80%, F1 score of 96.29%, and accuracy of 96.39%, respectively. Importantly, we scale out the verification system to imposter detection and achieve an average verification accuracy of 90.91%. These results imply the invariable characteristics from human factors rather than other traditional resources, which provides a superior solution for driving behavior authentication systems.

This study evaluated the factors influencing loyalty to an online clothing shop among college students. A survey was first conducted, and responses were compared. The main factors were identified and screened, and a sample of college students was selected. Likert-scale questionnaires were distributed, and the scores were calculated. The total score associated with the attitude of each respondent was ultimately determined. The strength of the respondents' attitude towards each topic was extracted and calculated, and the results were measured to determine the factors affecting loyalty to online clothing shops among college students. The top three influencing factors identified were as follows: the prices of the items sold by clothing stores, the fashion popularity of clothing, and clothing comfort. Compared with the female respondents, the male ones paid more attention to the matchability of clothing in stores and the fashion popularity of the goods sold. Compared with the males, the females paid more attention to clothing comfort, quality, the merchant's logistics speed and customer service attitude.

ABSTRACTMuscle development is a complex process that was regulated by many factors, among which non-coding RNAs (ncRNAs) play a vital role in regulating multiple life activities of muscle cells. Circular RNA (circRNA), a type of non-coding RNA with closed-loop structure, has been reported to affect multiple life processes. However, the roles of circRNAs on muscle development have not been fully elucidated. The present study aimed to determine whether and how circRIMKLBaffects muscle development. Our study revealed that circRIMKLBpromoted myoblast proliferation and inhibited differentiation. Besides, miR-29c was proved as a downstream target of circRIMKLBusing dual-luciferase reporter assay and RNA-binding protein immunoprecipitation (RIP) assay. Also, potassium inwardly rectifying channel subfamily J member 12 (KCNJ12) was identified as a novel target of miR-29cvia dual-luciferase reporter assay, quantitative reverse transcriptase-polymerase chain reaction (qRT-PCR), and western blot. CircRIMKLBand KCNJ12were both proved to regulate cell cycle on muscle regeneration after injury in vivo. In conclusion, we demonstrated that circRIMKLBsponged miR-29c, releasing KCNJ12to regulate myoblast proliferation and differentiation and regulating cell cycle during muscle regeneration after injury in vivo.

It is of great significance for safe driving to study drivers' eye movement and driving operation behavior when they encounter other road users violating traffic rules. The underlying reason is that most drivers are unable to process the unexpected visual stimulation, which is more likely to lead to driving accidents, especially in a hybrid situation. In this study, a driving simulator is used to design driving scenarios and study the driving performance of drivers with different driving experiences when other road users violate traffic rules. The experimental results show that some novice drivers ignore the position of their own vehicle when they encounter traffic violations which will lead to the collision with other road users. Moreover, some novice drivers can only perform one of the operations between steering and braking to avoid collision in these emergent situations. They cannot reasonably combine braking and steering to complete emergency driving operations like experienced drivers. Finally, when the driving difficulty increases, experienced drivers spend less time looking and more time scanning their surroundings to ensure that they can cope with the more complex driving environment while novice drivers do the opposite. This study reveals the difference between novices and experienced drivers, which paves a useful reference for the future advanced driving assistance system.