62The new coronavirus (SARS-CoV-2) outbreak originating from Wuhan, China, poses 63 a threat to global health. While it's evident that the virus invades respiratory tract and 64 transmits from human to human through airway, other viral tropisms and transmission 65
Similarity-preserving hashing is a widely-used method for nearest neighbour search in large-scale image retrieval tasks. For most existing hashing methods, an image is first encoded as a vector of hand-engineering visual features, followed by another separate projection or quantization step that generates binary codes. However, such visual feature vectors may not be optimally compatible with the coding process, thus producing sub-optimal hashing codes. In this paper, we propose a deep architecture for supervised hashing, in which images are mapped into binary codes via carefully designed deep neural networks. The pipeline of the proposed deep architecture consists of three building blocks: 1) a sub-network with a stack of convolution layers to produce the effective intermediate image features; 2) a divide-and-encode module to divide the intermediate image features into multiple branches, each encoded into one hash bit; and 3) a triplet ranking loss designed to characterize that one image is more similar to the second image than to the third one. Extensive evaluations on several benchmark image datasets show that the proposed simultaneous feature learning and hash coding pipeline brings substantial improvements over other state-of-the-art supervised or unsupervised hashing methods. * Corresponding author: Yan Pan, email: panyan5@mail.sysu.edu.cn. posed, e.g., [8,9,4,12,16,27,14,25,3]. The existing learning-based hashing methods can be categorized into unsupervised and supervised methods, based on whether supervised information (e.g., similarities or dissimilarities on data points) is involved. Compact bitwise representations are advantageous for improving the efficiency in both storage and search speed, particularly in big data applications. Compared to unsupervised methods, supervised methods usually embed the input data points into compact hash codes with fewer bits, with the help of supervised information.
The efficiency of perovskite solar cells has surged in the past few years, while the bandgaps of current perovskite materials for record efficiencies are much larger than the optimal value, which makes the efficiency far lower than the Shockley–Queisser efficiency limit. Here we show that utilizing the below-bandgap absorption of perovskite single crystals can narrow down their effective optical bandgap without changing the composition. Thin methylammonium lead triiodide single crystals with tuned thickness of tens of micrometers are directly grown on hole-transport-layer covered substrates by a hydrophobic interface confined lateral crystal growth method. The spectral response of the methylammonium lead triiodide single crystal solar cells is extended to 820 nm, 20 nm broader than the corresponding polycrystalline thin-film solar cells. The open-circuit voltage and fill factor are not sacrificed, resulting in an efficiency of 17.8% for single crystal perovskite solar cells.
Two-dimensional (2D) perovskites have been shown to be more stable than their three-dimensional (3D) counterparts due to the protection of the organic ligands. Herein a method is introduced to form 2D/3D stacking structures by the reaction of 3D perovskite with n-Butylamine (BA). Different from regular treatment with n-Butylammonium iodide (BAI) where 2D perovskite with various layers form, the reaction of BA with MAPbI only produce (BA)PbI, which has better protection due to more organic ligands in (BA)PbI than the mixture of 2D perovskites. Compared to BAI treatment, BA treatment results in smoother 2D perovskite layer on 3D perovskites with a better coverage. The photovoltaic devices with 2D/3D stacking structures show much improved stability in comparison to their 3D counterparts when subjected to heat stress tests. Moreover, the conversion of defective surface into 2D layers also induces passivation of the 3D perovskites resulting in an enhanced efficiency.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.