Amphibious Multifunctional Hydrogel Flexible Haptic Sensor with Self-Compensation Mechanism.

In recent years, hydrogel-based wearable flexible electronic devices have attracted much attention. However, hydrogel-based sensors are affected by structural fatigue, material aging, and water absorption and swelling, making stability and accuracy a major challenge. In this study, we present a DN-SPEZ dual-network hydrogel prepared using polyvinyl alcohol (PVA), sodium alginate (SA), ethylene glycol (EG), and ZnSO4 and propose a self-calibration compensation strategy. The strategy utilizes a metal salt solution to adjust the carrier concentration of the hydrogel to mitigate the resistance drift phenomenon to improve the stability and accuracy of hydrogel sensors in amphibious scenarios, such as land and water. The ExpGrow model was used to characterize the trend of the ∆R/R0 dynamic response curves of the hydrogels in the stress tests, and the average deviation of the fitted curves ϵ¯ was calculated to quantify the stability differences of different groups. The results showed that the stability of the uncompensated group was much lower than that of the compensated group utilizing LiCl, NaCl, KCl, MgCl2, and AlCl3 solutions (ϵ¯ in the uncompensated group in air was 276.158, 1.888, 2.971, 30.586, and 13.561 times higher than that of the compensated group in LiCl, NaCl, KCl, MgCl2, and AlCl3, respectively; ϵ¯ in the uncompensated group in seawater was 10.287 times, 1.008 times, 1.161 times, 4.986 times, 1.281 times, respectively, higher than that of the compensated group in LiCl, NaCl, KCl, MgCl2 and AlCl3). In addition, for the ranking of the compensation effect of different compensation solutions, the concentration of the compensation solution and the ionic radius and charge of the cation were found to be important factors in determining the compensation effect. Detection of events in amphibious environments such as swallowing, robotic arm grasping, Morse code, and finger-wrist bending was also performed in this study. This work provides a viable method for stability and accuracy enhancement of dual-network hydrogel sensors with strain and pressure sensing capabilities and offers solutions for sensor applications in both airborne and underwater amphibious environments.

Wavelength-Controlled Photoconductance Polarity Switching via Harnessing Defects in Doped PdSe2 for Artificial Synaptic Features.

Optoelectronic synapses are currently drawing significant attention as fundamental building blocks of neuromorphic computing to mimic brain functions. In this study, a two-terminal synaptic device based on a doped PdSe2 flake is proposed to imitate the key neural functions in an optical pathway. Due to the wavelength-dependent desorption of oxygen clusters near the intrinsic selenide vacancy defects, the doped PdSe2 photodetector achieves a high negative photoresponsivity of -7.8 × 103 A W-1 at 473 nm and a positive photoresponsivity of 181 A W-1 at 1064 nm. This wavelength-selective bi-direction photoresponse endows an all-optical pathway to imitate the fundamental functions of artificial synapses on a device level, such as psychological learning and forgetting capability, as well as dynamic logic functions. The underpinning photoresponse is further demonstrated on a flexible platform, providing a viable technology for neuromorphic computing in wearable electronics. Furthermore, the p-type doping results in an effective increase of the channel's electrical conductivity and a significant reduction in power consumption. Such low-power-consuming optical synapses with simple device architecture and low-dimensional features demonstrate tremendous promise for building multifunctional artificial neuromorphic systems in the future.

Size-fractionated N2 fixation off the Changjiang Estuary during summer.

Recent evidence has shown active N2 fixation in coastal eutrophic waters, yet the rate and controlling factors remain poorly understood, particularly in large estuaries. The Changjiang Estuary (CE) and adjacent shelf are characterized by fresh, nitrogen-replete Changjiang Diluted Water (CDW) and saline, nitrogen-depletion intruded Kuroshio water (Taiwan Warm Current and nearshore Kuroshio Branch Current), where N2 fixation may be contributed by different groups (i.e., Trichodesmium and heterotrophic diazotrophs). Here, for the first time, we provide direct measurement of size-fractionated N2 fixation rates (NFRs) off the CE during summer 2014 using the 15N2 bubble tracer method. The results demonstrated considerable spatial variations (southern > northern; offshore > inshore) in surface and depth-integrated NFRs, averaging 0.83 nmol N L-1 d-1 and 24.3 µmol N m-2 d-1, respectively. The highest bulk NFR (99.9 µmol N m-2 d-1; mostly contributed by >10 µm fraction) occurred in the southeastern East China Sea, where suffered from strong intrusion of the Kuroshio water characterized by low N/P ratio (<10) and abundant Trichodesmium (up to 10.23 × 106 trichomes m-2). However, low NFR (mostly contributed by <10 µm fraction) was detected in the CE controlled by the CDW, where NOx concentration (up to 80 µmol L-1) and N/P ratio (>100) were high and Trichodesmium abundance was low. The >10 µm fraction accounted for 60% of depth-integrated bulk NFR over the CE and adjacent shelf. We speculated that the present NFR of >10 µm fraction was mostly supported by Trichodesmium. Spearman rank correlation indicated that the NFR was significantly positively correlated with Trichodesmium abundance, salinity, temperature and Secchi depth, but was negatively with turbidity, N/P ratio, NOx, and chlorophyll a concentration. Our study suggests that distribution and size structure of N2 fixation off the CE are largely regulated by water mass (intruded Kuroshio water and CDW) movement and associated diazotrophs (particularly Trichodesmium) and nutrient conditions.

Application of explainable machine learning for real-time safety analysis toward a connected vehicle environment.

Due to the difficulty of obtaining traffic flow data and conflicts simultaneously, conflict-based analysis using macroscopic traffic features is much less studied. This research aims to analyze real-time safety by a disaggregate study and explore the benefit of the connected vehicle (CV) for real-time safety evaluation. To avoid the endogeneity problem regarding conflicts and traffic features in regression models, machine learning is employed to obtain a reliable and practical real-time safety model. The results show that the Random Forest outperforms eXtreme Gradient Boosting, Support Vector Machine and Adaptive Boosting models, achieving the best performance with the highest AUC of 0.827. For a deep understanding of conflict mechanisms, the explainable machine learning method SHAP (SHapley Additive exPlanation) is introduced to improve the model interpretability providing insights into the impacts of traffic flow features. Lane difference regarding average speed is found to have the most significant impacts on real-time safety. Speed variation, the proportion of trucks and traffic volume are associated with conflict occurrence. Further analysis highlights that the impacts of traffic features are heterogeneous and there may exist specific patterns of paired features affecting real-time safety. Encouragingly, SHAP appears to be able to complement the traditional model with random components in terms of revealing heterogeneity. The explainable machine learning can also provide a solid basis for discretizing continuous variables while previous studies perform discretization mainly based on prior knowledge and experience. The experimental result regarding CV Market Penetration Rate (CV-MPR) demonstrates that the model performance is gradually elevated with the increase of penetration rate. The initial stage of the CV market (20%, 40% CV-MPR) yields the most significant gains in real-time safety evaluation. These findings can be used beneficially in active traffic management.

Geometry Auxiliary Salient Object Detection for Light Fields via Graph Neural Networks.

Light field imaging, originated from the availability of light field capture technology, offers a wide range of applications in the field of computational vision. The capability of predicting salient objects of light fields remains technologically challenging due to its complicated geometry structure. In this paper, we propose a light field salient object detection approach that formulates the geometric coherence among multiple views of light fields as graphs, where the angular/central views represent the nodes and their relations compose the edges. The spatial and disparity correlations between multiple views are effectively explored through multi-scale graph neural networks, enabling the more comprehensive understanding of light field content and more representative and discriminative saliency features generation. Moreover, a multi-scale saliency feature consistency learning module is embedded to enhance the saliency features. Finally, an accurate salient object map is produced for the light field based upon the extracted features. In addition, we establish a new light field salient object detection dataset (CITYU-Lytro) that contains 817 light fields with diverse contents and their corresponding annotations, aiming to further promote the research on light field salient object detection. Quantitative and qualitative experiments demonstrate that the proposed method performs favorably compared with the state-of-the-art methods on the benchmark datasets.

Learning to Explore Saliency for Stereoscopic Videos via Component-Based Interaction.

In this paper, we devise a saliency prediction model for stereoscopic videos that learns to explore saliency inspired by the component-based interactions including spatial, temporal, as well as depth cues. The model first takes advantage of specific structure of 3D residual network (3D-ResNet) to model the saliency driven by spatio-temporal coherence from consecutive frames. Subsequently, the saliency inferred by implicit-depth is automatically derived based on the displacement correlation between left and right views by leveraging a deep convolutional network (ConvNet). Finally, a component-wise refinement network is devised to produce final saliency maps over time by aggregating saliency distributions obtained from multiple components. In order to further facilitate research towards stereoscopic video saliency, we create a new dataset including 175 stereoscopic video sequences with diverse content, as well as their dense eye fixation annotations. Extensive experiments support that our proposed model can achieve superior performance compared to the state-of-the-art methods on all publicly available eye fixation datasets.