Search

© 2020 Elsevier B.V. Facial expression recognition (FER) has a wide variety of applications ranging from human–computer interaction, robotics to health care. Although FER has made significant progress with the success of Convolutional Neural Network (CNN), it is still challenging especially for the video-based FER due to the dynamic changes in facial actions. Since the specific divergences exists among different expressions, we introduce a metric learning framework with a siamese cascaded structure that learns a fine-grained distinction for different expressions in video-based task. We also develop a pairwise sampling strategy for such metric learning framework. Furthermore, we propose a novel action-units attention mechanism tailored to FER task to extract spatial contexts from the emotion regions. This mechanism works as a sparse self-attention fashion to enable a single feature from any position to perceive features of the action-units (AUs) parts (eyebrows, eyes, nose, and mouth). Besides, an attentive pooling module is designed to select informative items over the video sequences by capturing the temporal importance. We conduct the experiments on four widely used datasets (CK+, Oulu-CASIA, MMI, and AffectNet), and also do experiment on the wild dataset AFEW to further investigate the robustness of our proposed method. Results demonstrate that our approach outperforms existing state-of-the-art methods. More in details, we give the ablation study of each component.

Lung nodule detection in chest X-ray (CXR) images is common to early screening of lung cancers. Deep-learning-based Computer-Assisted Diagnosis (CAD) systems can support radiologists for nodule screening in CXR images. However, it requires large-scale and diverse medical data with high-quality annotations to train such robust and accurate CADs. To alleviate the limited availability of such datasets, lung nodule synthesis methods are proposed for the sake of data augmentation. Nevertheless, previous methods lack the ability to generate nodules that are realistic with the shape/size attributes desired by the detector. To address this issue, we introduce a novel lung nodule synthesis framework in this paper, which decomposes nodule attributes into three main aspects including the shape, the size, and the texture, respectively. A GAN-based Shape Generator firstly models nodule shapes by generating diverse shape masks. The following Size Modulation then enables quantitative control on the diameters of the generated nodule shapes in pixel-level granularity. A coarse-to-fine gated convolutional Texture Generator finally synthesizes visually plausible nodule textures conditioned on the modulated shape masks. Moreover, we propose to synthesize nodule CXR images by controlling the disentangled nodule attributes for data augmentation, in order to better compensate for the nodules that are easily missed in the detection task. Our experiments demonstrate the enhanced image quality, diversity, and controllability of the proposed lung nodule synthesis framework. We also validate the effectiveness of our data augmentation strategy on greatly improving nodule detection performance.

Computer vision approaches have been widely used to automatically recognize the activities of workers from videos. While considerable advancements have been made to capture complementary information from still frames, it remains a challenge to obtain motion between them. As a result, this has hindered the ability to conduct real-time monitoring. Considering this challenge, an improved convolutional neural network (CNN) that integrates Red-Green-Blue (RGB), optical flow, and gray stream CNNs, is proposed to accurately monitor and automatically assess workers' activities associated with installing reinforcement during construction. A database containing photographs of workers installing reinforcement is created from activities undertaken on several construction projects in Wuhan, China. The database is then used to train and test the developed CNN network. Results demonstrate that the developed method can accurately detect the activities of workers. The developed computer vision-based approach can be used by construction managers as a mechanism to assist them to ensure that projects meet pre-determined deliverables.

Computer vision and pattern recognition approaches have been applied to determine unsafe behaviors on construction sites. Such approaches have been reliant on the computation of artificially complex image features that utilize a cumbersome parameter re-adjustment process. The creation of image features that can recognize unsafe actions, however, poses a significant research challenge on construction sites. This due to the prevailing complexity of spatio-temporal features, lighting, and the array of viewpoints that are required to identify an unsafe action. Considering these challenges, a new hybrid deep learning model that integrates a convolution neural network (CNN) and long short-term memory (LSTM) that automatically recognizes workers' unsafe actions is developed. The proposed hybrid deep learning model is used to: (1) identify unsafe actions; (2) collect motion data and site videos; (3) extract the visual features from videos using a CNN model; and (4) sequence the learning features that are enabled by the use of LSTM models. An experiment is used to test the model's ability to detect unsafe actions. The results reveal that the developed hybrid model (CNN + LSTM) is able to accurately detect safe/unsafe actions conducted by workers on-site. The model's accuracy exceeds the current state-of-the-art descriptor-based methods for detecting points of interest on images.

Ring finger protein 2 (RNF2), as a well-known E3 ligase, has an oncogenic role in various cancers. The role of RNF2 in colon cancer is still unknown. The aim of this work is to determine the biological role of RNF2 in colon cancer. We first examined the expression of RNF2 and interferon regulatory factor 4 (IRF4) in colon cancer patients and colon cancer cell lines (SW480 and HCT116). Compared with normal tumor-adjacent tissues, RNF2 was up-regulated whereas IRF4 was down-regulated in the colon cancer tissues. RNF2 was also up-regulated in colon cancer cells with respect to human fetal colon epithelial cells. RNF2 overexpression enhanced the ability of proliferation, migration and invasion of SW480 cells, whereas RNF2 knockdown caused an opposite result in HCT116 cells. Furthermore, a tumor xenograft model was constructed to verify the impact of RNF2 overexpressed-SW480 cells on tumor growth. RNF2 up-regulation elevated Ki-67 proliferation index, accelerated the growth of tumor tissues, and led to severe colon tissue damage in the tumor xenograft mice. In addition, RNF2 interacted with IRF4, and repressed IRF4 protein expression. IRF4 was a substrate of RNF2, and RNF2 promoted the ubiquitination and degradation of IRF4. RNF2 overexpression increased the ability of proliferation, migration and invasion in SW480 cells by promoting the ubiquitination and degradation of IRF4. In conclusion, this work demonstrated that RNF2 promoted tumor growth in colon cancer by regulating ubiquitination and degradation of IRF4. Thus, RNF2 may be served as a potential therapeutic target for colon cancer.

Penile erectile dysfunction (ED) caused by progressive fibrosis of the penis after bilateral cavernous nerve injury (BCNI) is one of the reasons for clinical treatment difficulties. Studies have shown that adipose-derived stem cells (ADSCs) could secrete abundant bioactive substances, which is an important reason for improving ED in BCNI rats. However, more information is needed to clarify what nutrient factors may play a leading role in preparing for effective clinical treatment in the future. The aim of this study was to investigate the effects of Lentivirus-mediated ADSCs overexpression of vascular endothelial growth factor (VEGF) and glial cells derived neurotrophic factors (GDNF) on cavernous nerve (CN) repair and penile erectile dysfunction. Two weeks after BCNI treatment, the CN was electrically stimulated and the erectile function of the rats was measured. Penile tissue and MPG were taken for blind histological analysis or western blotting. Erectile function recovered well in the ADSC-G&V group. ADSC-G&V group significantly increased the content of cavernous endothelial tissue and neuronal nitric oxide synthase, and improved the hypoxia of the corpus cavernosum. MPG from ADSC-G&V group showed morphological changes, including higher expression of NG-F and S100β. Two weeks later, PBS group displayed obviously fibrotic response, while the ADSC-G&V group was significantly inhibited and showed higher penile smooth muscle content and downregulated p-LIMK2/p-Cofilin pathway which was closely related to penile fibrosis after BCNI. Genetically modified ADSCs (GM-ADSCs) of VEGF and GDNF may promote rapid repair of CN and protect the local structure of the penis, thereby recovering erectile function better. Our experiments demonstrated that GM-ADSCs could be used as a means of rapid ameliorate of ED after BCNI, providing new ideas for future clinical treatment. Funding Statement: The study was supported by grants from the National Natural Science Foundation of China (No. 81971378); the Science and Technology Planning Project of Guangdong Province, China (2019B030316011); Sun Yat-Sen University 5010 Program (2015016). Declaration of Interests: The authors declare no conflict of interest. Ethics Approval Statement: All procedures were approved by the Ethics Committee of the Institutional Animal Care and Use Subcommittee of the Third Hospital of Sun Yat-sen University.

Objective The multi-wave longitudinal study was to explore the underlying mechanisms between sleep disturbance and depressive symptoms for up to 9.5 months among student nurses and to examine whether emotional exhaustion and sleep-related worry could be the mediators. Methods 856 student nurses aged 15–24 years (mean age = 17.8 years; all female) were recruited from a nursing school. Participants reported demographic variables (age, sibling, location, health condition, family monthly income and parents' marital quality) at baseline (Time1), Body Mass Index, weekly exercise, sleep disturbance and depressive symptoms at 3-month into internship (Time2), emotional exhaustion and sleep-related worry at 6-month into internship (Time3), and depressive symptoms at 9.5-month follow-up (Time4). Mediation analyses with bootstrapping were conducted to investigate the mediating role of emotional exhaustion and sleep-related worry in the relationship between sleep disturbance and depressive symptoms. Results Adjusted analyses suggested that the direct effect of sleep disturbance on depressive symptoms was not significant (β = 0.026, SE = 0.069, p = .71). Mediation analyses revealed two significant indirect effects between Time2 sleep disturbance and Time4 depressive symptoms with the mediating effect of emotional exhaustion (β = 0.019, SE = 0.009, BC95%CI [0.0036,0.0405]) and sleep-related worry (β = 0.016, SE = 0.008, BC95%CI [0.0026,0.0337]) respectively. When testing serial multiple mediation, the specific indirect effect of sleep disturbance on depressive symptoms through both emotional exhaustion and sleep-related worry was significant (β = 0.005, SE = 0.003, BC95%CI [0.0004,0.0123]). Conclusions Emotional exhaustion and sleep-related worry were positively associated with sleep disturbance and depressive symptoms.

The Aurivillius-type Bi2VO5.5-δ series doped by magnetic cations may have a significant electronic contribution to their total conductivity and special magnetoelectric properties which extend potential applications to magnetoelectronic and spintronic fields. Bi2(V1-xMnx)O5.5-δ and (Bi2-xMnx)VO5.5-δ (x = 0.10–0.30) compounds with nominal composition doped by manganese cation on the Bi- or V-site were prepared via a solid-state reaction. Results show that partial substitution of Mn on the V-site stabilizes Bi2(V1-xMnx)O5.5-δ in the high-temperature tetragonal structure of Bi2VO5.5-δ for compositions in the range 0.10 ≤ x

With the fast consumption of resource get and serious environmental pollution during electrical power's consumption, Efficiency Power Plant (EPP) concept is brought out. Although EPP gets more and more attentions, actual EPP's establishment is on a long way. This paper discusses theory and technology related with EPP. An actual instrument is produced to apply in distribution system, the measured electrical data in field shows that EPP is effective. At last, prospect of EPP is discussed.

Two stereoselective syntheses of SB-462795, a highly potent cathepsin K inhibitor, are described. Both routes feature a C5–C6 disconnection by ring closing metathesis to construct an azepane ring and are amenable to large-scale manufacturing.

Two types of water-insoluble protein from Bombyx mori silkworm, degummed and undegummed superfine silk powders, were prepared through a physical method. Water-soluble superfine silk fibroin powder was obtained from salt-dissolved, dialyzed and freeze-dried silk fibroin. These three types of superfine silk powders were blended with solution of commercially available biomedical grade segmented polyurethane (PU) to prepare elastic biofiber by a wet-spinning process. The objective of the current study was to identify the feasibility of wet spinning of polyurethane/superfine silk powder blend elastic biofiber, and analyse and compare the mechanical properties of the as-spun biofibers. Superfine silk fibroin powders and PU were dispersed in dimethyl formamide (DMF) and the dope solution exhibited good fiber formation in a warm distilled water coagulation bath. The superfine silk powder/PU blend fiber had a smooth surface, dense structure, and considerably high strength with a circular cross-section. The future further investigation will be carried out to analyse in detail the structure and morphology of the as-spun fiber. The silk-inspired biomedical polyurethane biofiber showed potential as a biomaterial for small-diameter artificial vascular scaffolds.

Polyurethane was blended with silk fibroin and heparin to prepare a heparin-releasing system. The release rate and the percentage of the cumulative amount of the released heparin can be controlled by the loading amount of heparin in the film, the composition ratio of silk fibroin to polyurethane, and the thickness of the film. The slower and more sustained release of heparin can be obtained by increasing the film thickness, the loading amount of heparin and the content of silk fibroin in the film. Thus the high bioactivity and long lasting antithrombogenicity of the raw heparin can be maintained for the blended film. The coagulation time tests showed that the composite film had good blood compatibility.