Search

Deep convolutional neural networks(DCNN) has recently significantly pushed the state of the art of semantic image segmentation. However, as the network depth increases, the progressively lower resolution leads to a loss of feature information and a large number of segmentation errors near the image edges. This is an imperative problem for applications that require accurate segmentation such as intrusion detection of railway. To address this problem, we propose superpixel pooling DCNN combined with random walks networks (SDR-Net), a semantic segmentation model that combines improved DeepLabV3+ and random walks algorithm. Firstly, the improved DeepLabV3+ with superpixel-pooling ASPP (SPASPP) is used as the basic structure of SDR-Net to introduce image prior information and improve the image feature extraction performance. Then, all pixels of the initial segmentation are classified by a binary classification network based on D2S transformation into two categories: correctly labeled pixels and incorrectly labeled pixels. Thirdly, the correctly labeled pixels are used as seed points, and then random walk algorithm is applied to reclassify the mislabeled pixels to achieve higher precision semantic image segmentation. Experiments on PASCAL VOC2012 datasets show the effectiveness of the algorithm in this paper, and it was successfully applied to rail intrusion detection.

The rapid development of industry intelligence has promoted the importance of sorting robots, but the shortcomings of sorting tasks are also gradually obvious. The target detection of the sorting robot in the process of sorting and grasping is affected by the anchor frame setting, and its grasping efficiency and recognition effect are low. Therefore, a detection network based on an anchor frame-free regression algorithm is proposed in this study, and a positioning and grasping method of the sorting robot based on the Deep Belief Network (DBN) algorithm is proposed. The algorithm experiment shows that the prediction accuracy of the proposed anchor-free detection network in the PASCAL VOC dataset reaches 92.91, significantly higher than that of other detection networks. In addition, the sorting robot was researched and designed with high accuracy in occlusion target recognition. When the occlusion area reaches 70%, the accuracy rate is as high as 81.26%, which is far higher than other detection schemes. The above results show that in improving the sorting robot’s target recognition ability, the anchor-free detection network can significantly improve its recognition accuracy. At the same time, introducing the DBN network can improve the robot’s recognition ability for occluded targets, which is of great significance to the development of the sorting robot and the intelligent industry. [ABSTRACT FROM AUTHOR]

Integrated electrochromic devices powered by photovoltaic cells have evoked a lot of interest due to their promising commercial prospects. However, their application has been restricted by the voltage adaption between the self-powered voltage and the color-changing threshold voltage (V t ). Herein, a strategy of bidirectional voltage regulating is proposed to develop a novel stand-alone integrated photovoltachromic device (I-PVCD), which integrates perovskite/organic tandem solar cells (P/O-TSCs) to drive color-changing process of conjugated poly(3-hexylthiophene) (P3HT) films. To lower the driving-voltage of electrochromic layer, C 60 is introduced to decrease the onset oxidation potential of P3HT film, and thus leading to a reduced V t of 0.70 V benefiting from the enhanced highest occupied molecular orbital level and decreased charge transfer resistance from 67.46 to 49.89 Ω. Simultaneously, PBDB-T is utilized as the hole transport layer in the interconnecting layer of CsPbI 2 Br/PTB7-Th:IEICO-4F P/O-TSC to improve its open-circuit voltage (V oc ) to 1.85 V. Under their synergetic merits, a I-PVCD with a wider self-adaptive voltage range is achieved. This device can undergo fast and reversible chromic transition from beautiful magenta to transparent only under the solar radiation, and demonstrates a coloration efficiency of 351.90 cm 2 C -1 and a switching time of 2 s besides its excellent operating reliability., (© 2024 Wiley‐VCH GmbH.)

Background: Against the backdrop of globalization, China remains one of the most heavily burdened countries in Asia with regard to AIDS. However, many high-risk groups and patients affected by AIDS may be less likely to actively seek care from medical institutions because of fear of experiencing shame or discrimination. Mobile apps provide a promising avenue for supporting the prevention, diagnosis, and treatment of AIDS. However, a comprehensive systematic evaluation of these mobile apps' functionality and quality has not been conducted yet., Objective: This study aims to identify the available mobile apps for AIDS in China, assess and discuss the functional features and quality of these Chinese AIDS mobile apps, and offer decision support for patients and clinical practitioners in accessing high-quality AIDS mobile apps. Furthermore, based on the evaluation results, recommendations for improvement will be provided., Methods: A systematic search was conducted on the Qimai app data platform, the Aladdin WeChat applet data platform, and WeChat to identify mobile apps related to AIDS. A snowball sampling method was used to supplement the potentially overlooked apps. The selected mobile apps underwent a rigorous screening process based on unified criteria. Subsequently, assessments were independently undertaken by 3 separate researchers and 2 patients with HIV, using both the Mobile App Rating Scale (MARS) and the User Mobile App Rating Scale (uMARS). Quantitative interpretations of the data were facilitated by the MedCalc statistical software (version 20.217, MedCalc Software)., Results: A total of 2901 AIDS mobile apps were included in the study, with 2897 identified through information retrieval and an additional 4 added via snowball sampling. After a rigorous selection process, 21 apps were determined to be usable. Among them, the Hong Feng Wan app achieved the highest combined average score, calculated based on the MARS (3.96, SD 0.33) and uMARS (4.47, SD 0.26). Overall, there was no significant correlation between MARS and uMARS (r app quality total score =0.41; P=.07; r subjective quality =0.39; P=.08). A notable issue was the widespread lack of user privacy protection, with only 24% (5/21) of the apps offering this feature., Conclusions: The number of available Chinese AIDS mobile apps is limited, with WeChat applets dominating the market. Nonetheless, the performance of WeChat mini-apps is generally inferior to that of independent apps, and there may be significant discrepancies between assessments conducted by researchers and those provided by genuine end users, emphasizing the necessity of involving real users in the development and evaluation of HIV mobile apps. In addition, developers of these Chinese HIV mobile apps need to devote attention to improving privacy protection mechanisms, in addition to considering the evaluations of researchers and real users. This will help attract more users and increase user loyalty., (©Peng Liu, Lingmeng Wang, Fuzhi Wang. Originally published in JMIR mHealth and uHealth (https://mhealth.jmir.org), 26.01.2024.)