Current methods for skeleton-based human action recognition usually work with complete skeletons. However, in real scenarios, it is inevitable to capture incomplete or noisy skeletons, which could ...significantly deteriorate the performance of current methods when some informative joints are occluded or disturbed. To improve the robustness of action recognition models, a multi-stream graph convolutional network (GCN) is proposed to explore sufficient discriminative features spreading over all skeleton joints, so that the distributed redundant representation reduces the sensitivity of the action models to non-standard skeletons. Concretely, the backbone GCN is extended by a series of ordered streams which is responsible for learning discriminative features from the joints less activated by preceding streams. Here, the activation degrees of skeleton joints of each GCN stream are measured by the class activation maps (CAM), and only the information from the unactivated joints will be passed to the next stream, by which rich features over all active joints are obtained. Thus, the proposed method is termed richly activated GCN (RA-GCN). Compared to the state-of-the-art (SOTA) methods, the RA-GCN achieves comparable performance on the standard NTU RGB+D 60 and 120 datasets. More crucially, on the synthetic occlusion and jittering datasets, the performance deterioration due to the occluded and disturbed joints can be significantly alleviated by utilizing the proposed RA-GCN.
Illustrations on the problems of current BBR losses. Each row shows the optimization results in different iterations with certain loss function. The Black denotes the anchor box. The Blue denotes the ...target box. The fist row denotes GIOU. The second row denotes CIOU. The third row denotes the proposed EIOU. EIOU attains more quick convergence speed and more accurate regression results. Display omitted
•We reveal the flaws of ℓn-norm and IOU-based losses for object detection.•We design a regression version of focal loss to emphasize the most promising anchors.•We conduct extensive experiments to validate the superiority of the proposed methods.
In object detection, bounding box regression (BBR) is a crucial step that determines the object localization performance. However, we find that most previous loss functions for BBR have two main drawbacks: (i) Both ℓn-norm and IOU-based loss functions are inefficient to depict the objective of BBR, which leads to slow convergence and inaccurate regression results. (ii) Most of the loss functions ignore the imbalance problem in BBR that the large number of anchor boxes which have small overlaps with the target boxes contribute most to the optimization of BBR. To mitigate the adverse effects caused thereby, we perform thorough studies to exploit the potential of BBR losses in this paper. Firstly, an Efficient Intersection over Union (EIOU) loss is proposed, which explicitly measures the discrepancies of three geometric factors in BBR, i.e., the overlap area, the central point and the side length. After that, we state the Effective Example Mining (EEM) problem and propose a regression version of focal loss to make the regression process focus on high-quality anchor boxes. Finally, the above two parts are combined to obtain a new loss function, namely Focal-EIOU loss. Extensive experiments on both synthetic and real datasets are performed. Notable superiorities on both the convergence speed and the localization accuracy can be achieved over other BBR losses.
Slot‐die coating is recognized as the most compatible method for the roll‐to‐roll (R2R) processing of large‐area flexible organic solar cells (OSCs). However, the photovoltaic performance of ...large‐area flexible OSC lags significantly behind that of traditional spin‐coating devices. In this work, two acceptors, Qx‐1 and Qx‐2, show quite different film‐formation kinetics in the slot‐die coating process. In situ absorption spectroscopy indicates that the excessive crystallinity of Qx‐2 provides early phase separation and early aggregation, resulting in oversized crystal domains. Consequently, the PM6:Qx‐1‐based 1 cm2 flexible device exhibits an excellent power conversion efficiency (PCE) of 13.70%, which is the best performance among the slot‐die‐coated flexible devices; in contrast, the PM6:Qx‐2 blend shows a pretty poor efficiency, which is lower than 1%. Moreover, the 30 cm2 modules based on PM6:Qx‐1, containing six 5 cm2 sub‐cells, exhibit a PCE of 12.20%. After being stored in a glove box for over 6000 h, the PCE remains at 103% of its initial values, indicating excellent shelf stability. Therefore, these results show a promising future strategy for the upscaling fabrication of flexible large‐area OSCs.
In situ absorption measurement is used to investigate the aggregation behavior of acceptors during slot‐die‐coating. The 1 cm2 flexible device can reach a power conversion efficiency of 13.70%, with excellent shelf stability and upscaling ability. The connected modules (180 cm2) can effectively power a smartphone, showing great potential for future applications.
Fiber Bragg grating (FBG) sensor has been considered as a reliable sensor for health monitoring of structural and geotechnical projects. Various types of FBG based sensors have been proposed in past ...few decades and employed for health monitoring of many geotechnical structures. This paper presents an overview of the recent development and application of FBG based sensors for health monitoring of several key geotechnical structures, including soil nail systems, slopes, and piles. Different sensor design, implementation and packaging methods, advantages and limitations of using FBG based sensors in different projects are reviewed. Comparative analysis of using two mathematical methods for the prediction of ground movement using FBG sensor data are also carried out. The two typical mathematical methods include Finite Difference Method (FDM) and Numerical Integration method (NIM). Possible technical challenges of applying FBG sensors for geotechnical monitoring are discussed.
One essential problem in skeleton-based action recognition is how to extract discriminative features over all skeleton joints. However, the complexity of the recent State-Of-The-Art (SOTA) models for ...this task tends to be exceedingly sophisticated and over-parameterized. The low efficiency in model training and inference has increased the validation costs of model architectures in large-scale datasets. To address the above issue, recent advanced separable convolutional layers are embedded into an early fused Multiple Input Branches (MIB) network, constructing an efficient Graph Convolutional Network (GCN) baseline for skeleton-based action recognition. In addition, based on such the baseline, we design a compound scaling strategy to expand the model's width and depth synchronously, and eventually obtain a family of efficient GCN baselines with high accuracies and small amounts of trainable parameters, termed EfficientGCN-Bx, where "x" denotes the scaling coefficient. On two large-scale datasets, i.e. , NTU RGB+D 60 and 120, the proposed EfficientGCN-B4 baseline outperforms other SOTA methods, e.g. , achieving 92.1% accuracy on the cross-subject benchmark of NTU 60 dataset, while being 5.82× smaller and 5.85× faster than MS-G3D, which is one of the SOTA methods. The source code in PyTorch version and the pretrained models are available at https://github.com/yfsong0709/EfficientGCNv1 .
Background
Microsurgical bone reconstruction is challenging in cases of anterior maxillary defects because of the presence of residual teeth. The study analyzed the method and feasibility of using ...vascularized bone flaps to reconstruct anterior maxillary defects.
Methods
We retrospectively analyzed 15 patients treated at the Department of Oral and Maxillofacial Surgery, Peking University School, and Hospital of Stomatology between November 2017 and January 2023.
Results
A total of 14 patients were successfully reconstructed using the digital technique: 6 deep circumflex iliac artery (DCIA) flaps and 8 fibular free flaps. Palatal and buccal pedicle paths were used in 4 and 10 cases, respectively, for vascular anastomosis. The implant coverage rates of the DCIA and fibular free flaps were 87.2% and 92.6%, respectively (p > 0.05).
Conclusion
Anterior maxillary defects reconstructed with the DCIA and fibular free flaps are reliable methods. Furthermore, the palatal and buccal paths are recommended for vascular anastomosis.
Spark plasma sintering (SPS), also known as pulsed electric current sintering (PECS) or field-assisted sintering technique (FAST), belongs to a class of powder metallurgy techniques. In SPS, the ...sample is simultaneously subjected to a uniaxial pressure and electrical current in a vacuum or protective atmosphere. Although the fundamental principles of this procedure were first proposed over 50 years ago, SPS acquired major importance only within the last 20 years. Scholars come to realize that SPS technique enables control of the powder surface condition, atomic diffusion behavior, and phase stability and crystal growth behavior, as well as accelerating densification of hard-to-sinter materials. This review summarizes the latest research findings with respect to experimental procedures, densification behaviors, microstructural characteristics, and mechanical properties of various traditional and novel materials synthesized using SPS, mainly highlighting the heating mechanisms in SPS and the effects induced by multi-physical fields on materials. In addition, influences of operating parameters containing current, voltage, and uniaxial pressure on product characteristics are reviewed for a wide range of materialsincluding hard-to-sinter materials, carbon-containing materials, nanocrystalline materials, non-equilibrium materials, gradient materials, interconnect materials, complex shape materials, and advanced functional materials.
Display omitted
•Effects of current, electric fields and pressure on SPS processing and micro-mechanisms in materials are investigated.•The state-of-the-art concepts of SPS devices for R&D and industrial application are presented.•The common technological superiority and optimization mechanisms of applying SPS in every material system are illustrated.•Overview on the design, experiment and simulation of both structural and functional materials developed by SPS technique.
With reduced background and high sensitivity, photoelectrochemistry (PEC) may be applied as an intracellular nanotool and open a new technological direction of single‐cell study. Nevertheless, the ...present palette of single‐cell tools lacks such a PEC‐oriented solution. Here a dual‐functional photocathodic single‐cell nanotool capable of direct electroosmotic intracellular drug delivery and evaluation of oxidative stress is devised by engineering a target‐specific organic molecule/NiO/Ni film at the tip of a nanopipette. Specifically, the organic molecule probe serves simultaneously as the biorecognition element and sensitizer to synergize with p‐type NiO. Upon intracellular delivery at picoliter level, the oxidative stress effect will cause structural change of the organic probe, switching its optical absorption and altering the cathodic response. This work has revealed the potential of PEC single‐cell nanotool and extended the boundary of current single‐cell electroanalysis.
An integrated photocathodic nanotool was fabricated for dual‐functional intracellular drug delivery and evaluation of cellular oxidative stress in single live cell.
Titanium metal–organic frameworks (Ti‐MOFs), as an appealing type of artificial photocatalyst, have shown great potential in the field of solar energy conversion due to their well‐studied photoredox ...activity (similar to TiO2) and good optical responsiveness of linkers, which serve as the antenna to absorb visible‐light. Although much effort has been dedicated to developing Ti‐MOFs with high photocatalytic activity, their solar energy conversion performances are still poor. Herein, we have implemented a covalent‐integration strategy to construct a series of multivariate Ti‐MOF/COF hybrid materials PdTCPP⊂PCN‐415(NH2)/TpPa (composites 1, 2, and 3), featuring excellent visible‐light utilization, a suitable band gap, and high surface area for photocatalytic H2 production. Notably, the resulting composites demonstrated remarkably enhanced visible‐light‐driven photocatalytic H2 evolution performance, especially for the composite 2 with a maximum H2 evolution rate of 13.98 mmol g−1 h−1 (turnover frequency (TOF)=227 h−1), which is much higher than that of PdTCPP⊂PCN‐415(NH2) (0.21 mmol g−1 h−1) and TpPa (6.51 mmol g−1 h−1). Our work thereby suggests a new approach to highly efficient photocatalysts for H2 evolution and beyond.
A series of covalently connected multivariate Ti‐MOF/COF hybrid materials were constructed demonstrating outstanding photocatalytic H2 evolution performance with a maximum H2 evolution rate of 13.98 mmol g−1 h−1 (TOF=227 h−1), much higher than the prototypical counterparts.
Organic mechanoluminochromic materials are mechano/piezo‐responsive and promising for applications in sensors, displays, and data storage devices. However, their switching range of emission is ...seriously impeded by only one kind of emission (either a fluorescent or phosphorescent peak) in the spectrum of single organic compounds. This study presents a design strategy for pure organic compounds with excellent room‐temperature fluorescent–phosphorescent dual‐emission (rFPDE) properties, which combines the effective factors of dipenylsulfone group, crystalline state, and heavy atom effect. Following the principle of color mixing, myriad emission colors with a wide range from orange to purple and across white zone in a straight line in the chromaticity diagram of the Commission Internationale de l’Eclairage (CIE) can be obtained by simply mechanical grinding the compound. The unique properties could be concentrated on a pure organic compound through this design strategy, which provides a new efficient channel for the discovery of efficient mechano‐responsive organic materials.
Grinding a rainbow: A purely organic compound with excellent room‐temperature fluorescent–phosphorescent dual‐emission (rFPDE) properties brought about by mechanical grinding has been demonstrated. It shows linearly tunable emission colors with a straight line across the white zone in the chromaticity diagram, following the principle of color mixing.