Seeking a Hierarchical Prototype for Multimodal Gesture Recognition.

Gesture recognition has drawn considerable attention from many researchers owing to its wide range of applications. Although significant progress has been made in this field, previous works always focus on how to distinguish between different gesture classes, ignoring the influence of inner-class divergence caused by gesture-irrelevant factors. Meanwhile, for multimodal gesture recognition, feature or score fusion in the final stage is a general choice to combine the information of different modalities. Consequently, the gesture-relevant features in different modalities may be redundant, whereas the complementarity of modalities is not exploited sufficiently. To handle these problems, we propose a hierarchical gesture prototype framework to highlight gesture-relevant features such as poses and motions in this article. This framework consists of a sample-level prototype and a modal-level prototype. The sample-level gesture prototype is established with the structure of a memory bank, which avoids the distraction of gesture-irrelevant factors in each sample, such as the illumination, background, and the performers' appearances. Then the modal-level prototype is obtained via a generative adversarial network (GAN)-based subnetwork, in which the modal-invariant features are extracted and pulled together. Meanwhile, the modal-specific attribute features are used to synthesize the feature of other modalities, and the circulation of modality information helps to leverage their complementarity. Extensive experiments on three widely used gesture datasets demonstrate that our method is effective to highlight gesture-relevant features and can outperform the state-of-the-art methods.

Dual-Affinity Style Embedding Network for Semantic-Aligned Image Style Transfer.

Image style transfer aims at synthesizing an image with the content from one image and the style from another. User studies have revealed that the semantic correspondence between style and content greatly affects subjective perception of style transfer results. While current studies have made great progress in improving the visual quality of stylized images, most methods directly transfer global style statistics without considering semantic alignment. Current semantic style transfer approaches still work in an iterative optimization fashion, which is impractically computationally expensive. Addressing these issues, we introduce a novel dual-affinity style embedding network (DaseNet) to synthesize images with style aligned at semantic region granularity. In the dual-affinity module, feature correlation and semantic correspondence between content and style images are modeled jointly for embedding local style patterns according to semantic distribution. Furthermore, the semantic-weighted style loss and the region-consistency loss are introduced to ensure semantic alignment and content preservation. With the end-to-end network architecture, DaseNet can well balance visual quality and inference efficiency for semantic style transfer. Experimental results on different scene categories have demonstrated the effectiveness of the proposed method.

Vitamin D ameliorates high-fat-diet-induced hepatic injury via inhibiting pyroptosis and alters gut microbiota in rats.

Accumulating evidence suggests that vitamin D (VD) has a therapeutic effect on non-alcoholic fatty liver disease (NAFLD). Pyroptosis and gut microbiota have been recognized as critical factors of the progression of NAFLD. However, the effect of VD on the pyroptosis and gut microbiota in NAFLD remains inconclusive. Herein, rats were fed high fat diet (HFD) for 12 weeks and concurrently treated with 5 µg/kg 1,25(OH)2D3 twice a week. BRL-3A cells were stimulated with 0.4 mmol/L palmitic acid (PA) and 1 µg/ml lipopolysaccharide (LPS) for 16 h and treated with 10-6 mol/L 1,25(OH)2D3. Effect of VD on the hepatic injury, lipid accumulation, activation of NLRP3 inflammasome and pyroptosis was determined in vivo and in vitro. Next, gasdermin D N-terminal (GSDMD-N) fragment was overexpressed in BRL-3A cells to investigate the role of pyroptosis in the therapeutic effect of VD on NAFLD. In addition, gut microbiota in NAFLD rats was also analyzed. Results showed that VD attenuated HFD-induced hepatic injury in vivo and PA-LPS-induced impairment of cell viability in vitro, and inhibited lipid accumulation, activation of NLRP3 inflammasome and pyroptosis in vivo and in vitro. GSDMD-N fragment overexpression suppressed the protective effect of VD on PA-LPS-induced activation of NLRP3 inflammasome, impairment of cell viability and lipid accumulation, indicating that VD might attenuate NAFLD through inhibiting pyroptosis. Additionally, VD also restored HFD-induced gut microbiota dysbiosis by increasing the relative abundance of Lactobacillus and reducing that of Acetatifactor, Oscillibacter and Flavonifractor. This study provides a novel mechanism underlying VD therapy against NAFLD.

Bioinformatic Analysis of Crosstalk Between circRNA, miRNA, and Target Gene Network in NAFLD.

Background: The majority of chronic liver disease is caused by non-alcoholic fatty liver disease (NAFLD), which is one of the highly prevalent diseases worldwide. The current studies have found that non-coding RNA (ncRNA) plays an important role in the NAFLD, but few studies on circRNA. In this study, genes, microRNA (miRNA), and circular RNA (circRNA) associated with NAFLD were found by bioinformatic methods, bringing a novel perspective for the prevention and treatment of NAFLD. Methods: Expression data of GSE63067 was acquired from Gene Expression Omnibus (GEO) database. The liver samples were collected from the people diagnosed with NAFLD or not. Differentially expressed genes (DEGs) were obtained from the steatosis vs. the control group and non-alcoholic steatohepatitis (NASH) vs. the control group using the GEO2R online tool. The overlapping genes remained for further functional enrichment analysis and protein-protein interaction network analysis. MiRNAs and circRNAs targeting these overlapping DEGs were predicted from the databases. Finally, the GSE134146 dataset was used to verify the expression of circRNA. Results: In summary, 228 upregulated and 63 downregulated differential genes were selected. The top 10 biological processes and relative signaling pathways of the upregulated differential genes were obtained. Also, ten hub genes were performed in the Protein-protein interaction (PPI) network. One hundred thirty-nine miRNAs and 902 circRNAs were forecast for the differential genes by the database. Ultimately, the crosstalk between hsa_circ_0000313, miR-6512-3p, and PEG10 was constructed. Conclusion: The crosstalk of hsa_circ_0000313-hsa-miR-6512-3p-PEG10 and some related non-coding RNAs may take part in NAFLD's pathogenesis, which could be the potential biomarkers of NAFLD in the future.

Clostridium butyricum Alleviates Gut Microbiota Alteration-Induced Bone Loss after Bariatric Surgery by Promoting Bone Autophagy.

Bariatric surgery is the most common and effective treatment of severe obesity; however, these bariatric procedures always result in detrimental effects on bone metabolism by underlying mechanisms. This study aims to investigate the skeletal response to bariatric surgery and to explore whether Clostridium butyricum alleviates gut microbiota alteration-induced bone loss after bariatric surgery. Consequently, male SD rats received Roux-en-Y gastric bypass (RYGB) and sleeve gastrectomy (SG) surgery, respectively, followed by body weight recording. The bone loss after bariatric surgery was further determined by dual-energy X-ray absorptiometry (DXA), micro-CT measurement, histologic analyses, and Western blot. Besides, 16S rDNA gene sequencing was performed to determine the gut microbiota alteration after surgery, and intervention with fecal microbiota from RYGB donor was conducted in obese SD rats, followed by C. butyricum administration. Accordingly, rats in the RYGB and SG groups maintained sustained weight loss, and DXA and micro-CT measurement further demonstrated significant bone loss after bariatric surgery. Besides, histologic and Western blot analyses validated enhanced osteoclastogenesis and inhibited osteoblastogenesis and defective autophagy after surgery. The 16S rDNA gene sequencing suggested a significant alteration of gut microbiota composition in the RYGB group, and intervention with fecal microbiota from RYGB donor further determined that this kind of alteration contributed to the bone loss after RYGB. Meanwhile, C. butyricum might protect against this postoperative bone loss by promoting osteoblast autophagy. In summary, this study suggests novel mechanisms to clarify the skeletal response to bariatric surgery and provides a potential candidate for the treatment of bone disorder among bariatric patients. SIGNIFICANCE STATEMENT: The significance of this study is the discovery of obvious bone loss and defective autophagy after bariatric surgery. Besides, it is revealed that gut microbiota alterations could be the reason for impaired bone mass after bariatric surgery. Furthermore, Clostridium butyricum could alleviate the gut microbiota alteration-induced bone loss after bariatric surgery by promoting osteoblast autophagy.

Liraglutide ameliorates lipotoxicity-induced inflammation through the mTORC1 signalling pathway.

Lipotoxicity has been implicated in many disease processes, and prolonged exposure to high lipid levels often leads to the activation of a variety of abnormal signals, which in turn leads to the induction of inflammation. The aim of our study was to explore the correlation between mammalian target of rapamycin (mTOR) and inflammation by studying high-fat diet (HFD)-induced non-alcoholic fatty liver disease (NAFLD) in rats and palmitate (PA)-induced inflammation (lipotoxicity) in HepG2 cells. In addition, we investigated whether the glucagon-like peptide-1 (GLP-1) analogue liraglutide can protect rats and HepG2 cells from lipotoxicity. Our results showed that an HFD and PA significantly increased inflammation by activating the mTORC1 pathway in vitro and in vivo. Treatment with rapamycin (an mTOR inhibitor) inhibited some effects of PA on inflammation. Furthermore, we observed that liraglutide inhibited PA-induced inflammation by inactivating mTORC1 signalling molecules. Overall, our findings demonstrated that mTORC1 signalling pathways were involved primarily in high lipid level-induced inflammation. Importantly, liraglutide may protect against lipotoxicity-induced inflammation by regulating mTORC1-dependent pathways.

Image style transfer with collection representation space and semantic-guided reconstruction.

Image style transfer renders the content of an image into different styles. Current methods made decent progress with transferring the style of single image, however, visual statistics from one image cannot reflect the full scope of an artist. Also, previous work did not put content preservation in the important position, which would result in poor structure integrity, thus deteriorating the comprehensibility of generated image. These two problems would limit the visual quality improvement of style transfer results. Targeting at style resemblance and content preservation problems, we propose a style transfer system composed of collection representation space and semantic-guided reconstruction. We train an encoder-decoder network with art collections to construct a representation space that can reflect the style of the artist. Then, we use semantic information as guidance to reconstruct the target representation of the input image for better content preservation. We conduct both quantitative analysis and qualitative evaluation to assess the proposed method. Experiment results demonstrate that our approach well balanced the trade-off between capturing artistic characteristics and preserving content information in style transfer tasks.