Nanjing University of Science and Technology

About: Nanjing University of Science and Technology is a education organization based out in Nanjing, China. It is known for research contribution in the topics: Catalysis & Computer science. The organization has 31581 authors who have published 36390 publications receiving 525474 citations. The organization is also known as: Nánjīng Lǐgōng Dàxué & Nánlǐgōng.

Pyramid Vision Transformer: A Versatile Backbone for Dense Prediction without Convolutions

Abstract: Although using convolutional neural networks (CNNs) as backbones achieves great successes in computer vision, this work investigates a simple backbone network useful for many dense prediction tasks without convolutions. Unlike the recently-proposed Transformer model (e.g., ViT) that is specially designed for image classification, we propose Pyramid Vision Transformer~(PVT), which overcomes the difficulties of porting Transformer to various dense prediction tasks. PVT has several merits compared to prior arts. (1) Different from ViT that typically has low-resolution outputs and high computational and memory cost, PVT can be not only trained on dense partitions of the image to achieve high output resolution, which is important for dense predictions but also using a progressive shrinking pyramid to reduce computations of large feature maps. (2) PVT inherits the advantages from both CNN and Transformer, making it a unified backbone in various vision tasks without convolutions by simply replacing CNN backbones. (3) We validate PVT by conducting extensive experiments, showing that it boosts the performance of many downstream tasks, e.g., object detection, semantic, and instance segmentation. For example, with a comparable number of parameters, RetinaNet+PVT achieves 40.4 AP on the COCO dataset, surpassing RetinNet+ResNet50 (36.3 AP) by 4.1 absolute AP. We hope PVT could serve as an alternative and useful backbone for pixel-level predictions and facilitate future researches. Code is available at this https URL.

Semiconducting Group 15 Monolayers: A Broad Range of Band Gaps and High Carrier Mobilities

Abstract: Optoelectronic applications require materials both responsive to objective photons and able to transfer carriers, so new two-dimensional (2D) semiconductors with appropriate band gaps and high mobilities are highly desired. A broad range of band gaps and high mobilities of a 2D semiconductor family, composed of monolayer of Group 15 elements (phosphorene, arsenene, antimonene, bismuthene) is presented. The calculated binding energies and phonon band dispersions of 2D Group 15 allotropes exhibit thermodynamic stability. The energy band gaps of 2D semiconducting Group 15 monolayers cover a wide range from 0.36 to 2.62 eV, which are crucial for broadband photoresponse. Significantly, phosphorene, arsenene, and bismuthene possess carrier mobilities as high as several thousand cm2 V−1 s−1. Combining such broad band gaps and superior carrier mobilities, 2D Group 15 monolayers are promising candidates for nanoelectronics and optoelectronics.

Two-dimensional antimonene single crystals grown by van der Waals epitaxy.

Abstract: Unlike the unstable black phosphorous, another two-dimensional group-VA material, antimonene, was recently predicted to exhibit good stability and remarkable physical properties. However, the synthesis of high-quality monolayer or few-layer antimonenes, sparsely reported, has greatly hindered the development of this new field. Here, we report the van der Waals epitaxy growth of few-layer antimonene monocrystalline polygons, their atomical microstructure and stability in ambient condition. The high-quality, few-layer antimonene monocrystalline polygons can be synthesized on various substrates, including flexible ones, via van der Waals epitaxy growth. Raman spectroscopy and transmission electron microscopy reveal that the obtained antimonene polygons have buckled rhombohedral atomic structure, consistent with the theoretically predicted most stable β-phase allotrope. The very high stability of antimonenes was observed after aging in air for 30 days. First-principle and molecular dynamics simulation results confirmed that compared with phosphorene, antimonene is less likely to be oxidized and possesses higher thermodynamic stability in oxygen atmosphere at room temperature. Moreover, antimonene polygons show high electrical conductivity up to 104 S m−1 and good optical transparency in the visible light range, promising in transparent conductive electrode applications. Several two-dimensional materials have been synthesized to date, yet elemental materials, consisting of individual atomic species, are still scarce. Here, the authors synthesize few-layer, monocrystalline polygons of antimonene via van der Waals epitaxy growth.

Richer Convolutional Features for Edge Detection

Abstract: Edge detection is a fundamental problem in computer vision. Recently, convolutional neural networks (CNNs) have pushed forward this field significantly. Existing methods which adopt specific layers of deep CNNs may fail to capture complex data structures caused by variations of scales and aspect ratios. In this paper, we propose an accurate edge detector using richer convolutional features (RCF). RCF encapsulates all convolutional features into more discriminative representation, which makes good usage of rich feature hierarchies, and is amenable to training via backpropagation. RCF fully exploits multiscale and multilevel information of objects to perform the image-to-image prediction holistically. Using VGG16 network, we achieve state-of-the-art performance on several available datasets. When evaluating on the well-known BSDS500 benchmark, we achieve ODS F-measure of 0.811 while retaining a fast speed (8 FPS). Besides, our fast version of RCF achieves ODS F-measure of 0.806 with 30 FPS. We also demonstrate the versatility of the proposed method by applying RCF edges for classical image segmentation.

Making strong nanomaterials ductile with gradients

Abstract: Steels can be made stronger, tougher, or more resistant to corrosion either by changing composition (adding in more carbon or other elements) or by modifying their microstructures. An extreme microstructural route for strengthening materials is to reduce the crystallite size from the micrometer scale (“coarse-grained”) to the nanoscale. Nanograined aluminum or copper (Cu) may become even harder than high-strength steels, but these materials can be very brittle and crack when pulled (deformed in tension), apparently because strain becomes localized and resists deformation. However, nanograined metals can be plastically deformed under compression or rolling at ambient temperature, implying that moderate deformation can occur if the cracking process is suppressed. Tremendous efforts have been made to explore how to suppress strain localization in tensioned nanomaterials and make them ductile. Gradient microstructures, in which the grain size increases from nanoscale at the surface to coarse-grained in the core, were recently discovered to be an effective approach to improving ductility ( 1 – 4 ).