Metric networks for enhanced perception of non-local semantic information.

Introduction: Metric learning, as a fundamental research direction in the field of computer vision, has played a crucial role in image matching. Traditional metric learning methods aim at constructing two-branch siamese neural networks to address the challenge of image matching, but they often overlook to cross-source and cross-view scenarios. Methods: In this article, a multi-branch metric learning model is proposed to address these limitations. The main contributions of this work are as follows: Firstly, we design a multi-branch siamese network model that enhances measurement reliability through information compensation among data points. Secondly, we construct a non-local information perception and fusion model, which accurately distinguishes positive and negative samples by fusing information at different scales. Thirdly, we enhance the model by integrating semantic information and establish an information consistency mapping between multiple branches, thereby improving the robustness in cross-source and cross-view scenarios. Results: Experimental tests which demonstrate the effectiveness of the proposed method are carried out under various conditions, including homologous, heterogeneous, multi-view, and crossview scenarios. Compared to the state-of-the-art comparison algorithms, our proposed algorithm achieves an improvement of ~1, 2, 1, and 1% in terms of similarity measurement Recall@10, respectively, under these four conditions. Discussion: In addition, our work provides an idea for improving the crossscene application ability of UAV positioning and navigation algorithm.

Insights Into to the KX (X = Cl, Br, I) Adsorption-Assisted Stabilization of CsPbI2 Br Surface.

Despite the excellent optoelectronic properties, organic-inorganic hybrid perovskite solar cells (PSCs) still present significant challenges in terms of ambient stability. CsPbI2 Br, a member of all-inorganic perovskites, may respond to this challenge because of its inherent high stability against light, moisture, and heat, and therefore has gained tremendous attraction recently. However, the practical application of CsPbI2 Br is still impeded by the notorious phenomenon of photoinduced halide segregation. Herein, by applying first-principles calculations, the stability, electronic structure, defect properties, and ion-diffusion properties of the stoichiometric CsPbI2 Br (110) surface and that with the adsorption of KX (X = Cl, Br, I) are systematically investigated. It is found that the adsorbed KX can serve as an external substitute of the halogen vacancies on the surface, therefore inhibiting halogen segregation and improving the stability of the CsPbI2 Br surface. The KX can also eliminate deep-level defect states caused by antisites, thereby contributing to the promoted optoelectronic properties of CsPbI2 Br. The mechanistic understanding of surface passivation in this work can lay the foundation for the future design of CsPbI2 Br PSCs with optimized optoelectronic performance.

Rare postoperative hemorrhage after robotic-assisted pancreatoduodenectomy for pancreatic head cancer: a case report.

Post-pancreaticoduodenectomy hemorrhage is a life-threatening complication that occurs in 2-10% of patients. The most common location for post-pancreaticoduodenectomy hemorrhage is the gastroduodenal artery stump. Nonetheless, unusual sources of hemorrhage, which are hard to locate, exist. Here, we report a rare postoperative hemorrhage after robotic-assisted pancreatoduodenectomy for pancreatic head cancer. A 67-year-old man presenting with appetite loss, general fatigue and painless jaundice was admitted to our ward. The patient had an elevated level of carbohydrate antigen 19-9 (50 U/mL). Computed tomography scan revealed a 17-mm wide low-density area in the uncinate process of the pancreas. Magnetic resonance cholangiopancreatography showed the dilation of bile and pancreatic ducts. Robotic-assisted pancreaticoduodenectomy was performed on the patient by using the da Vinci Model S Surgical System. On postoperative days 5 and 6, the patient vomited blood, and bloody fluid was observed in the drainage. Emergent gastroscopic examination was performed and revealed a large amount of hematocele in the stomach. On postoperative day 6, emergency operation was undertaken, and the output jejunal loop was found to have intussuscepted in the stomach. This is the first case report of output jejunal loop intussusception in the stomach that consequently caused hemorrhage after robotic-assisted pancreaticoduodenectomy for pancreatic head cancer.

Green Emission Induced by Intrinsic Defects in All-Inorganic Perovskite CsPb2Br5.

All-inorganic perovskites with improved stability are expected to be better candidates for optoelectronics, compared to organic-inorganic hybrid perovskites. A new member of all-inorganic perovskites, CsPb2Br5, has attracted great attention for its promising applications in optoelectronic devices. However, the origins of the green emission in CsPb2Br5 have been actively debated. By using first-principles calculations, we find that CsPb and VBr are dominant intrinsic defects independent of the growth conditions within the stable region of CsPb2Br5. Interestingly, we suggest that individual intrinsic defects do not lead to the green emission of CsPb2Br5, while the donor-acceptor pair recombination of CsPb and VBr possibly does. Our findings provide new insights into the experimental controversy about the green emission and its origins in CsPb2Br5 from the perspective of intrinsic defects, which help to extend the application of CsPb2Br5 in optoelectronic devices.

Security of Semi-Device-Independent Random Number Expansion Protocols.

Semi-device-independent random number expansion (SDI-RNE) protocols require some truly random numbers to generate fresh ones, with making no assumptions on the internal working of quantum devices except for the dimension of the Hilbert space. The generated randomness is certified by non-classical correlation in the prepare-and-measure test. Until now, the analytical relations between the amount of the generated randomness and the degree of non-classical correlation, which are crucial for evaluating the security of SDI-RNE protocols, are not clear under both the ideal condition and the practical one. In the paper, first, we give the analytical relation between the above two factors under the ideal condition. As well, we derive the analytical relation under the practical conditions, where devices' behavior is not independent and identical in each round and there exists deviation in estimating the non-classical behavior of devices. Furthermore, we choose a different randomness extractor (i.e., two-universal random function) and give the security proof.

[Effects of shugan jianpi gusui recipe on multiple sclerosis recurrence: a primary report].

OBJECTIVE: To observe the effects of Shugan Jianpi Gusui Recipe (SJGR) on multiple sclerosis (MS). METHODS: A case cohort study was used. The MS patients were assigned to the test group (14 cases) and the control group (21 cases) according to whether they would combine with Chinese herbs. Corticosteroids or gamma globulin was administered to all patients in the acute phase. SJGR was administered to patients in the test group in the remission phase, while those in the control group were not treated or treated by azathioprine alone. They were followed-up for a long time after their first visits. The observation time ranged 10-131 successive months. The recurrence intervals and the yearly average recurrence times were calculated in the two groups. RESULTS: When compared with before treatment, the recurrence interval was obviously prolonged, and the yearly average recurrence times decreased in the test group after treatment with statistical difference (P < 0.05). There was statistical difference in the recurrence interval and the yearly average recurrence times between the test group and the control group (P < 0.05). CONCLUSIONS: SJGR showed better effects in prolonging the recurrence interval and reducing the yearly average recurrence times of MS patients. It is worth further researches.

ZEB2 promotes the metastasis of gastric cancer and modulates epithelial mesenchymal transition of gastric cancer cells.

BACKGROUND: Invasion and metastasis are the hallmarks of advanced gastric cancer progression. Therefore, it is urgent to overcome metastasis in order to improve the survival of gastric cancer patients. AIMS: This study aimed to examine the expression of ZEB2 in gastric cancer samples and analyze its correlation with clinicopathologic features. In addition, the molecular mechanism by which ZEB2 contributes to gastric cancer metastasis will be explored. METHODS: ZEB2 expression in clinical gastric cancer samples was evaluated by immunohistochemical analysis. ZEB2 was knocked-down in HGC27 gastric cancer cells by shRNA and the effects on cell invasion and migration were examined by in vitro cell invasion and migration assays. The expression of epithelial marker E-cadherin, mesenchymal markers fibronecin and vimentin, and MMPs was detected by western blot analysis. RESULTS: The expression of ZEB2 was positively correlated with the depth of invasion, lymph node metastasis and TNM stage. In addition, patients with positive ZEB2 expression showed a significantly shorter overall survival time than did patients with negative ZEB2. shRNA mediated knockdown of ZEB2 resulted in reduced invasion and migration of HGC27 cells, along with the upregulation of E-cadherin and downregulation of fibronecin, vimentin, MMP2, and MMP9. CONCLUSIONS: ZEB2 expression is closely associated with the clinicopathological parameters of gastric cancer. ZEB2 promotes gastric cancer cell migration and invasion at least partly via the regulation of epithelial-mesenchymal transition. ZEB2 is a potential target for gene therapy of aggressive gastric cancer.