Temporal Modulation Research Articles

3D‐Printed In Situ Growth of Bilayer MOF Hydrogels for Accelerated Osteochondral Defect Repair

AbstractRepairing osteochondral (OC) defect presents a significant challenge due to the intricate structural requirements and the unpredictable differentiation pathways of bone marrow mesenchymal stem cells (BMSCs). To address this challenge, a novel biomimetic OC hydrogel scaffold is developed that features a structure of soft and hard components. This scaffold incorporates bilayer metal–organic frameworks (MOFs), specifically ZIF‐67 in the upper layer and ZIF‐8 in the lower layer, achieved through an in situ printing process. This configuration enables the spatial and temporal modulation of BMSC differentiation by controlling the release of Co2⁺ and Zn2⁺. The results demonstrate that the bilayer MOF hydrogels significantly outperform hydrogels that either lack MOFs or contain a single type of MOF in enhancing repair outcomes in rabbit models of knee OC defects. The improved regenerative efficacy is attributed to the distinct chondrogenic and osteogenic differentiation cues provided by the bilayer MOFs, effectively guiding BMSCs toward enhanced tissue regeneration. This customizable biomimetic OC hydrogel scaffold not only opens new avenues for innovative therapeutic strategies but also holds great promise for widespread clinical applications.

Partial prior transfer learning based on self-attention CNN for EEG decoding in stroke patients.

The utilization of motor imagery-based brain-computer interfaces (MI-BCI) has been shown to assist stroke patients activate motor regions in the brain. In particular, the brain regions activated by unilateral upper limb multi-task are more extensive, which is more beneficial for rehabilitation, but it also increases the difficulty of decoding. In this paper, self-attention convolutional neural network based partial prior transfer learning (SACNN-PPTL) is proposed to improve the classification performance of patients' MI multi-task. The backbone network of the algorithm is SACNN, which accords with the inherent features of electroencephalogram (EEG) and contains the temporal feature module, the spatial feature module and the feature generalization module. In addition, PPTL is introduced to transfer part of the target domain while preserving the generalization of the base model while improving the specificity of the target domain. In the experiment, five backbone networks and three training modes are selected as comparison algorithms. The experimental results show that SACNN-PPTL had a classification accuracy of 55.4%±0.17 in four types of MI tasks for 22 patients, which is significantly higher than comparison algorithms (P < 0.05). SACNN-PPTL effectively improves the decoding performance of MI tasks and promotes the development of BCI-based rehabilitation for unilateral upper limb.

Mechanisms of Increasing Weld Depth during Temporal Power Modulation in High‐Power Laser Beam Welding

Understanding the fundamental mechanisms of action for increasing weld depth during temporal power modulation in laser beam welding could allow dissimilar rotational welding without the introduction of concomitant turbulence, but with enhanced intermixing. The investigations are conducted on 30 mm‐diameter round bars of stainless steel alloy 1.4301 and nickel base alloy 2.4856 utilizing a 16 kW disk laser beam source. Modulation frequencies are 0/50/100/200 Hz at low, medium, and high amplitudes of laser beam power. The influence on the process and weld characteristics is investigated through high‐speed imaging with grayscale analysis, keyhole depth measurements, metallographic sections, and energy‐dispersive X‐ray spectroscopy analysis. The objectives are successfully achieved, and the underlying mechanism is maintaining the keyhole depth at a higher level for modulation frequencies of 200 Hz and a high amplitude of laser beam power, which is related to the keyhole inertia. Based on this, a novel welding mode with a constant keyhole depth is proposed. Furthermore, up to 20% increase in weld depth is achieved, a saturation limit for the modulation frequency is identified, intermixing within the weld is enhanced, and a model for predicting the weld depth based solely on measurements of the surface width is developed.

SAM-Net: Spatio-Temporal Sequence Typhoon Cloud Image Prediction Net with Self-Attention Memory

Cloud image prediction is a spatio-temporal sequence prediction task, similar to video prediction. Spatio-temporal sequence prediction involves learning from historical data and using the learned features to generate future images. In this process, the changes in time and space are crucial for spatio-temporal sequence prediction models. However, most models now rely on stacking convolutional layers to obtain local spatial features. In response to the complex changes in cloud position and shape in cloud images, the prediction module of the model needs to be able to extract both global and local spatial features from the cloud images. In addition, for irregular cloud motion, more attention should be paid to the spatio-temporal sequence features between input cloud image frames in the temporal sequence prediction module, considering the extraction of temporal features with long temporal dependencies, so that the spatio-temporal sequence prediction network can learn cloud motion trends more accurately. To address these issues, we have introduced an innovative model called SAM-Net. The self-attention module of this model aims to extract an inner image frame’s spatial features of global and local dependencies. In addition, a memory mechanism has been added to the self-attention module to extract interframe features with long temporal and spatial dependencies. Our method shows better performance than the PredRNN-v2 model on publicly available datasets such as MovingMNIST and KTH. We achieved the best performance in both the 4-time-step and 10-time-step typhoon cloud image predictions. On a cloud dataset consisting of 10 time steps, we observed a decrease in MSE of 180.58, a decrease in LPIPS of 0.064, an increase in SSIM of 0.351, and a significant improvement in PSNR of 5.56 compared to PredRNN-v2.

Optically tuned soliton dynamics in Bose-Einstein condensates within dark traps

Abstract This study investigates the formation and dynamics of solitons in Bose–Einstein condensates (BECs) within dark traps generated by two crossed Laguerre–Gaussian (LG) beams with varying azimuthal indices ℓ . As the index ℓ increases, the potential transitions from a harmonic trap when ℓ = 1 to a square-well potential for larger values of ℓ . This transition allows us to study a range of soliton dynamics under different confinement conditions while maintaining the same BEC volume. Through the derivation of the Gross-Pitaevskii equation (GPE) and under these specific conditions in both one-dimensional and two-dimensional configurations, we explore the dynamics of solitons across multiple scenarios. The study examines two primary methods for solitons generation: the temporal modulation of the scattering length and the implementation of an initial potential barrier that is subsequently removed. The results indicate that the trap shape plays a critical role in the generation and interaction dynamics of solitons. In harmonic traps, solitons exhibit a behavior different from those observed in anharmonic traps, where the dynamics is significantly influenced by the azimuthal index of the trap. The ability to control soliton dynamics in BECs holds significant promise for applications in quantum technologies, precision sensing, and the exploration of fundamental quantum phenomena.

Binocular integration of chromatic and luminance signals.

Much progress has been made in understanding how the brain combines signals from the two eyes. However, most of this work has involved achromatic (black and white) stimuli, and it is not clear if the same processes apply in color-sensitive pathways. In our first experiment, we measured contrast discrimination ("dipper") functions for four key ocular configurations (monocular, binocular, half-binocular, and dichoptic), for achromatic, isoluminant L-M and isoluminant S-(L+M) sine-wave grating stimuli (L: long-, M: medium-, S: short-wavelength). We find a similar pattern of results across stimuli, implying equivalently strong interocular suppression within each pathway. Our second experiment measured dichoptic masking within and between pathways using the method of constant stimuli. Masking was strongest within-pathway and weakest between S-(L+M) and achromatic mechanisms. Finally, we repeated the dipper experiment using temporal luminance modulations, which produced slightly weaker interocular suppression than for spatially modulated stimuli. We interpret our results in the context of a contemporary two-stage model of binocular contrast gain control, implemented here using a hierarchical Bayesian framework. Posterior distributions of the weight of interocular suppression overlapped with a value of 1 for all dipper data sets, and the model captured well the pattern of thresholds from all three experiments.

Optimizing HMG-CoA Synthase Expression for Enhanced Limonene Production in Escherichia coli through Temporal Transcription Modulation Using Optogenetics.

Overexpression of a single enzyme in a multigene heterologous pathway may be out of balance with the other enzymes in the pathway, leading to accumulated toxic intermediates, imbalanced carbon flux, reduced productivity of the pathway, or an inhibited growth phenotype. Therefore, optimal, balanced, and synchronized expression levels of enzymes in a particular metabolic pathway is critical to maximize production of desired compounds while maintaining cell fitness in a growing culture. Furthermore, the optimal intracellular concentration of an enzyme is determined by the expression strength, specific timing/duration, and degradation rate of the enzyme. Here, we modulated the intracellular concentration of a key enzyme, namely HMG-CoA synthase (HMGS), in the heterologous mevalonate pathway by tuning its expression level and period of transcription to enhance limonene production in Escherichia coli. Facilitated by the tuned blue-light inducible BLADE/pBad system, we observed that limonene production was highest (160 mg/L) with an intermediate transcription level of HMGS from moderate light illumination (41 au, 150 s ON/150 s OFF) throughout the growth. Owing to the easy penetration and removal of blue-light illumination from the growing culture which is hard to obtain using conventional chemical-based induction, we further explored different induction patterns of HMGS under strong light illumination (2047 au, 300 s ON) for different durations along the growth phases. We identified a specific timing of HMGS expression in the log phase (3-9 h) that led to optimal limonene production (200 mg/L). This is further supported by a mathematical model that predicts several periods of blue-light illumination (3-9 h, 0-9 h, 3-12 h, 0-12 h) to achieve an optimal expression level of HMGS that maximizes limonene production and maintains cell fitness. Compared to moderate and prolonged transcription (41 au, 150 s ON/150 s OFF, 0-73 h), strong but time-limited transcription (2047 au, 300 s ON, 3-9 h) of HMGS could maintain its optimal intracellular concentration and further increased limonene production up to 92% (250 mg/L) in the longer incubation (up to 73 h) without impacting cell fitness. This work has provided new insight into the "right amount" and "just-in-time" expression of a critical metabolite enzyme in the upper module of the mevalonate pathway using optogenetics. This study would complement previous findings in modulating HMGS expression and potentially be applicable to heterologous production of other terpenoids in E. coli.

Temporal modulation (early escalation and late de-escalation) of antiplatelet therapy in patients undergoing complex high-risk PCI: rationale and design of the TAILORED-CHIP trial.

Despite the use of conventional dual antiplatelet therapy (DAPT) after percutaneous coronary intervention (PCI), the risk of adverse events remains high among patients with increased thrombotic risk. Until recently, the optimal antiplatelet strategy to balance the ischaemic and bleeding risks in patients who are undergoing complex high-risk PCI has been unclear. The TAILored Versus COnventional AntithRombotic StratEgy IntenDed for Complex HIgh-Risk PCI (TAILORED-CHIP) trial is an investigator-initiated, multicentre, prospective randomised trial to evaluate the efficacy and safety of a time-dependent tailored antiplatelet therapy with an early (<6 months post-PCI) escalation (low-dose ticagrelor at 60 mg twice daily plus aspirin) and a late (>6 months post-PCI) de-escalation (clopidogrel monotherapy) in patients undergoing complex high-risk PCI as compared with standard DAPT (clopidogrel plus aspirin for 12 months). Eligible patients had to have at least one high-risk anatomical or procedural feature or clinical characteristic associated with an increased risk of ischaemic or thrombotic events. The primary endpoint was the net clinical outcome, a composite of death from any cause, myocardial infarction, stroke, stent thrombosis, urgent revascularisation, or clinically relevant bleeding (Bleeding Academic Research Consortium type 2, 3, or 5) at 12 months after randomisation. (ClinicalTrials.gov: NCT03465644).

TEA-GCN: Transformer-Enhanced Adaptive Graph Convolutional Network for Traffic Flow Forecasting.

Traffic flow forecasting is crucial for improving urban traffic management and reducing resource consumption. Accurate traffic conditions prediction requires capturing the complex spatial-temporal dependencies inherent in traffic data. Traditional spatial-temporal graph modeling methods often rely on fixed road network structures, failing to account for the dynamic spatial correlations that vary over time. To address this, we propose a Transformer-Enhanced Adaptive Graph Convolutional Network (TEA-GCN) that alternately learns temporal and spatial correlations in traffic data layer-by-layer. Specifically, we design an adaptive graph convolutional module to dynamically capture implicit road dependencies at different time levels and a local-global temporal attention module to simultaneously capture long-term and short-term temporal dependencies. Experimental results on two public traffic datasets demonstrate the effectiveness of the proposed model compared to other state-of-the-art traffic flow prediction methods.

Laser power modulation between disturbed and undisturbed material removal regime in laser chemical processing

Laser chemical machining (LCM) is a method for removing material by thermally induced chemical dissolution of self-passivating metals. However, the process window is limited by disrupted material removal due to gas bubble formation and metallic salts and oxides deposition at higher energy input. Since the temperature increases, and therefore gas bubble growth takes time, it is hypothesized that the temporal modulation of laser power can remove the metal homogeneously, i.e., without disrupted material removal, while achieving a higher removal rate. Based on this, the dynamic process behavior of material removal is investigated for the LCM of titanium in phosphoric acid, using a rectangular modulation of the laser power with varying irradiation durations. As a result, however, high-speed videos show that gas bubbles are consistently generated, regardless of the applied laser power and power modulation, although the quantity of bubbles varies with different parameters. Even with short power durations (10 ms), the material deposition occurs after multiple irradiations. When the duration is longer, the material deposition increases in height along the laser scan direction. For the studied process parameters, a Fourier analysis in the spatial domain further indicates the correlation between the material removal frequencies and the modulation frequencies. In conclusion, the laser power modulation cannot prevent the disturbed material removal at high laser powers. Nevertheless, the material deposition can be utilized to generate periodic surface structures with a depth below and above the initial surface.

Transfer of solitons and half-vortex solitons via adiabatic passage

We show that the transfer of matter-wave solitons and half-vortex solitons in a spin-orbit coupled Bose-Einstein condensate between two (or more) arbitrarily chosen sites of an optical lattice can be implemented using the adiabatic passage. The underlying linear Hamiltonian has a flat band in its spectrum, so that even sufficiently weak interatomic interactions can sustain well-localized Wannier solitons which are involved in the transfer process. The adiabatic passage is assisted by properly chosen spatial and temporal modulations of the Rabi frequency. Within the framework of a few-mode approximation, the mechanism is enabled by a dark state created by coupling the initial and target low-energy solitons with a high-energy extended Bloch state, like in the conventional stimulated Raman adiabatic passage used for the coherent control of quantum states. In real space, however, the atomic transfer between initial and target states is sustained by the current carried by the extended Bloch state, which remains populated during the whole process. The full description of the transfer is provided by the Gross-Pitaevskii equation. Protocols for the adiabatic passage are described for one- and two-dimensional optical lattices, as well as for splitting and subsequent transfer of an initial wave packet simultaneously to two different target locations. Published by the American Physical Society 2024

Forced flow reversal in ferrofluidic Couette flow via alternating magnetic field

Time-dependent boundary conditions are very common in natural and industrial flows and by far no exception. An example of this is the movement of a magnetic fluid forced due to temporal modulations. In this study, we used numerical methods to examine the dynamics of ferrofluidic wavy vortex flows (WVF2, with dominant azimuthal wavenumber m=2) in the counter-rotating Taylor–Couette system, which was subjected to time-periodic modulation/forcing in a spatially homogeneous magnetic field. In the absence of a magnetic field, all WVF2 states move in the opposite direction to the rotation of the inner cylinder, they are retrograde. However, when strength or frequency of the alternating magnetic field increases, the motion direction of the flow pattern changes. Thus, the alternating field provides a precise and controllable key parameter for triggering the system response and controlling the flow. Aside, we also observed intermittent behavior when one solution became unstable, leading to random transitions in both, the transition time and toward the different final solutions. Our findings suggest that, in ferrofluids, flow pattern reversal can be induced by varying a magnetic field in a controlled manner, which may have applications in the development of modern fluid devices in laboratory experiments. These findings provide a framework to study other types of magnetic flows driven by time-dependent forcing.

Harnessing the Natural Resonances of Time-Varying Dispersive Interfaces.

Space-time modulation of electromagnetic parameters offers novel exciting possibilities for advanced field manipulations. In this study, we explore wave scattering from a time-varying interface characterized by a Lorentz-type dispersion with a steplike temporal variation in its parameters. Our findings reveal a new process: an unconventional frequency generation at the natural resonances of the system. Remarkably, this phenomenon enables the coupling of propagating waves to evanescent ones, allowing the direct far-field excitation of surface-wave modes without the mediation of spatial gratings or prisms. These results suggest a novel strategy for designing compact and ultrafast photonic devices, eliminating the necessity for subwavelength spatial structuring or prolonged temporal modulations.

Extended multi-stream temporal-attention module for skeleton-based human action recognition (HAR)

Graph convolutional networks (GCNs) are an effective skeleton-based human action recognition (HAR) technique. GCNs enable the specification of CNNs to a non-Euclidean frame that is more flexible. The previous GCN-based models still have a lot of issues: (I) The graph structure is the same for all model layers and input data. GCN model's hierarchical structure and human action recognition input diversity make this a problematic approach; (II) Bone length and orientation are understudied due to their significance and variance in HAR. For this purpose, we introduce an Extended Multi-stream Temporal-attention Adaptive GCN (EMS-TAGCN). By training the network topology of the proposed model either consistently or independently according to the input data, this data-based technique makes graphs more flexible and faster to adapt to a new dataset. A spatial, temporal, and channel attention module helps the adaptive graph convolutional layer focus on joints, frames, and features. Hence, a multi-stream framework representing bones, joints, and their motion enhances recognition accuracy. Our proposed model outperforms the NTU RGBD for CS and CV by 0.6% and 1.4%, respectively, while Kinetics-skeleton Top-1 and Top-5 are 1.4% improved, UCF-101 has improved 2.34% accuracy and HMDB-51 dataset has significantly improved 1.8% accuracy. According to the results, our model has performed better than the other models. Our model consistently outperformed other models, and the results were statistically significant that demonstrating the superiority of our model for the task of HAR and its ability to provide the most reliable and accurate results.

Algorithm Improvement for Mobile Event Detection with Intelligent Tunnel Robots

Mobile inspections conducted by intelligent tunnel robots are instrumental in broadening the inspection reach, economizing on inspection expenditures, and augmenting the operational efficiency of inspections. Despite differences from fixed surveillance, mobile-captured traffic videos have complex backgrounds and device conditions that interfere with accurate traffic event identification, warranting more research. This paper proposes an improved algorithm based on YOLOv9 and DeepSORT for intelligent event detection in an edge computing mobile device using an intelligent tunnel robot. The enhancements comprise the integration of the Temporal Shift Module to boost temporal feature recognition and the establishment of logical rules for identifying diverse traffic incidents in mobile video imagery. Experimental results show that our fused algorithm achieves a 93.25% accuracy rate, an improvement of 1.75% over the baseline. The algorithm is also applicable to inspection vehicles, drones, and autonomous vehicles, effectively enhancing the detection of traffic events and improving traffic safety.

Multi-Attention Recurrent Neural Network for Multi-Step Prediction of Chlorophyll Concentration

Abstract: Chlorophyll prediction facilitates the comprehension of red tide characteristics and enables early warning. In practice, it is formulated as a multivariate time series forecasting problem aimed at forecasting future chlorophyll concentrations by considering both exogenous factors and chlorophyll. However, the multi-step prediction of chlorophyll concentration poses a formidable challenge due to the intricate interaction between factors and the long temporal dependence between input sequences. In this work, we propose a Multi-attention Recurrent Neural Network (MaRNN) for the multi-step prediction of chlorophyll concentration. The MaRNN comprises an encoder incorporating two-stage spatial attention and a decoder employing temporal attention. The encoder first learns the significance of exogenous factors for prediction in the first phase, and subsequently captures the spatial correlation between the exogenous sequence and chlorophyll sequence in the second phase. The decoder further excavates input sequences that exhibit a strong correlation with the task through temporal attention module, thereby enhancing the prediction accuracy of the model. Experiments conducted on two real-world datasets reveal that MaRNN not only surpasses state-of-the-art methods in performance, but also offers interpretability for chlorophyll prediction.

Characteristic of temporal light modulation reference meter and standard light source for calibration and verification

Abstract The measurement of the temporal light modulation (TLM) of a light source is gaining increasing importance owing to its detrimental effects on human vision and health. To establish traceability in TLM measurements, the National Institute of Metrology, China (NIM), established a TLM reference meter (TLMM) and standard TLM light source (STLS) for calibration and verification services. The TLMM and STLS employ detector- and source-based approaches, respectively, to define the standards for TLM quantities. The TLMM rapidly acquires light waveform data from the TLM light sources and calculates corresponding TLM quantities. Frequency calibration result is with an expanded uncertainty of 1 × 10−5, frequency response function is almost flat in range of 1 kHz–25 kHz, and response time is less than 8 × 10−6 s. The electrical noise was evaluated at approximately 1000:1–5000:1. The STLS generates an arbitrary TLM waveform by converting the virtual waveform into an actual LED output light waveform. For the frequency response function, the relative amplitude decreased by 1% with a frequency of 15 kHz, and response time was less than 5 × 10−6 s. The linearity of the frequency dependency was studied with ratio varied by approximately 0.19% when the periodic triangle excitation in frequency band from 1 Hz to 1 kHz. Experimental comparisons demonstrated that TLM quantities (frequency, percentage flicker, flicker index, short-term flicker severity P st LM , and stroboscopic effect visibility measure ( SVM )) are consistent between the TLMM and STLS devices on 14 typical TLM waveforms. For P st LM , most of the deviations were in the range of −0.019–0.079 (except for one with 0.182). For SVM , most of the deviations were in the range of −0.001–0.003 (except for one with 0.014).

Spatially and Temporally Optimized Audio‐Driven Talking Face Generation

AbstractAudio‐driven talking face generation is essentially a cross‐modal mapping from audio to video frames. The main challenge lies in the intricate one‐to‐many mapping, which affects lip sync accuracy. And the loss of facial details during image reconstruction often results in visual artifacts in the generated video. To overcome these challenges, this paper proposes to enhance the quality of generated talking faces with a new spatio‐temporal consistency. Specifically, the temporal consistency is achieved through consecutive frames of the each phoneme, which form temporal modules that exhibit similar lip appearance changes. This allows for adaptive adjustment in the lip movement for accurate sync. The spatial consistency pertains to the uniform distribution of textures within local regions, which form spatial modules and regulate the texture distribution in the generator. This yields fine details in the reconstructed facial images. Extensive experiments show that our method can generate more natural talking faces than previous state‐of‐the‐art methods in both accurate lip sync and realistic facial details.

Visibility and annoyance of the phantom array effect varies with age and history of migraine

The phantom array effect (PAE) is a series of repeated images that may be perceived when a person moves their eyes in large saccades across a light source (or a specular reflection of that light source) that is modulating in output over time. Fifty-five people, including a group of 25 who experience migraine, evaluated the visibility and annoyingness of phantom arrays produced by 85 unique temporal light modulation waveforms (including sine, rectangular, complex and DC waveforms) generated using an LED placed against a black background. Those with migraine exhibited higher average visibility compared to those without migraine ( p = 0.019) and were relatively more sensitive at higher frequencies ( p &lt; 0.001). Younger participants also found more stimuli to be visible ( p &lt; 0.001). The threshold sensitivity function was similar to that developed for the phantom array visibility measure (PAVM), and PAVM was effective in predicting visibility ( R2 = 0.87 for the relevant region of PAVM &lt; 3). While those in the migraine group did not report seeing the PAE more often in everyday life at a statistically significant level, they reported being more annoyed by it and having more unwanted physiological responses (headaches, eye fatigue and distraction/disorientation). Members of the migraine group were also more likely to have changed their behaviour in architectural spaces (such as leaving a restaurant with ‘flickering’ lights). In the four hours after completing the experiment, 64% of the migraine group (vs. 19% of the non-migraine group) reported experiencing discomfort or an adverse reaction. In particular, 41% reported experiencing a headache (vs. 8% for those in the non-migraine group).

Improvement of temporal light artefact effects in commercial LED lamps

Temporal light modulation and its potential adverse visual effects on human observers, collectively known as temporal light artefacts (TLAs), have been limited by the European Union Ecodesign regulation. Limitations on two key TLA metrics, flicker and stroboscopic effect, were introduced in September 2021. This study analyses the impact of this regulation on commercial LED lamps by measuring multiple different consumer-grade LED E27 retrofit lamps by several manufacturers before and after the regulation. The lamps purchased after the Ecodesign regulation seem to favour a certain type of LED driver that provides the best TLA behaviour of all lamp types. The regulation, with respect to TLA, has resulted in the replacement of ill-behaving lamp types in the market by newer lamps with better TLA characteristics.