A Model for Infrastructure Detection along Highways Based on Remote Sensing Images from UAVs.

Infrastructure along the highway refers to various facilities and equipment: bridges, culverts, traffic signs, guardrails, etc. New technologies such as artificial intelligence, big data, and the Internet of Things are driving the digital transformation of highway infrastructure towards the future goal of intelligent roads. Drones have emerged as a promising application area of intelligent technology in this field. They can help achieve fast and precise detection, classification, and localization of infrastructure along highways, which can significantly enhance efficiency and ease the burden on road management staff. As the infrastructure along the road is exposed to the outdoors for a long time, it is easily damaged and obscured by objects such as sand and rocks; on the other hand, based on the high resolution of the images taken by Unmanned Aerial Vehicles (UAVs), the variable shooting angles, complex backgrounds, and high percentage of small targets mean the direct use of existing target detection models cannot meet the requirements of practical applications in industry. In addition, there is a lack of large and comprehensive image datasets of infrastructure along highways from UAVs. Based on this, a multi-classification infrastructure detection model combining multi-scale feature fusion and an attention mechanism is proposed. In this paper, the backbone network of the CenterNet model is replaced with ResNet50, and the improved feature fusion part enables the model to generate fine-grained features to improve the detection of small targets; furthermore, the attention mechanism is added to make the network focus more on valuable regions with higher attention weights. As there is no publicly available dataset of infrastructure along highways captured by UAVs, we filter and manually annotate the laboratory-captured highway dataset to generate a highway infrastructure dataset. The experimental results show that the model has a mean Average Precision (mAP) of 86.7%, an improvement of 3.1 percentage points over the baseline model, and the new model performs significantly better than other detection models overall.

R-YOLO: A YOLO-Based Method for Arbitrary-Oriented Target Detection in High-Resolution Remote Sensing Images.

In view of the existence of remote sensing images with large variations in spatial resolution, small and dense objects, and the inability to determine the direction of motion, all these components make object detection from remote sensing images very challenging. In this paper, we propose a single-stage detection network based on YOLOv5. This method introduces the MS Transformer module at the end of the feature extraction network of the original network to enhance the feature extraction capability of the network model and integrates the Convolutional Block Attention Model (CBAM) to find the attention area in dense scenes. In addition, the YOLOv5 target detection network is improved by incorporating a rotation angle approach from the a priori frame design and the bounding box regression formulation to make it suitable for rotating frame-based detection scenarios. Finally, the weighted combination of the two difficult sample mining methods is used to improve the focal loss function, so as to improve the detection accuracy. The average accuracy of the test results of the improved algorithm on the DOTA data set is 77.01%, which is higher than the previous detection algorithm. Compared with the average detection accuracy of YOLOv5, the average detection accuracy is improved by 8.83%. The experimental results show that the algorithm has higher detection accuracy than other algorithms in remote sensing scenes.

CGR-Block: Correlated Feature Extractor and Geometric Feature Fusion for Point Cloud Analysis.

Point cloud processing based on deep learning is developing rapidly. However, previous networks failed to simultaneously extract inter-feature interaction and geometric information. In this paper, we propose a novel point cloud analysis module, CGR-block, which mainly uses two units to learn point cloud features: correlated feature extractor and geometric feature fusion. CGR-block provides an efficient method for extracting geometric pattern tokens and deep information interaction of point features on disordered 3D point clouds. In addition, we also introduce a residual mapping branch inside each CGR-block module for the further improvement of the network performance. We construct our classification and segmentation network with CGR-block as the basic module to extract features hierarchically from the original point cloud. The overall accuracy of our network on the ModelNet40 and ScanObjectNN benchmarks achieves 94.1% and 83.5%, respectively, and the instance mIoU on the ShapeNet-Part benchmark also achieves 85.5%, proving the superiority of our method.

Regenerable multifunctional mesoporous silica nanocomposites for simultaneous detection and removal of mercury(II).

Mercury (Hg(2+)) is a highly toxic and widespread environmental pollutant. Herein, a regenerable and highly selective core-shell structured magnetic mesoporous silica nanocomposite with functionalization of thymine (T) and T-rich DNA (denoted as Fe3O4@nSiO2@mSiO2-T-TRDNA nanocomposite) has been developed for simultaneous detection and removal of Hg(2+). In this work, the thymine and T-rich DNA were immobilized onto the interior and exterior surface of outermost mesoporous silica, respectively. The detection mechanism is based on Hg(2+)-mediated hairpin structure formed by T-rich DNA functionalized on the exterior surface of the nanocomposites, where, upon addition of SYBR Green I dye, strong fluorescence is observed. In the absence of Hg(2+), however, addition of the dye results in low fluorescence. The limit of detection for Hg(2+) in a buffer is 2 nM by fluorescence spectroscopy. Simultaneously, the Fe3O4@nSiO2@mSiO2-T-TRDNA nanocomposite features a selective binding with Hg(2+) between two thymines immobilized at the interior surface of the mesopores and exhibits efficient and convenient Hg(2+) removal by a magnet. Kinetic study reveals that the Hg(2+) removal is a rapid process with over 80% of Hg(2+) removed within approximately 1 h. The applicability of the developed nanocomposites is demonstrated to detect and remove Hg(2+) from samples of Xiangjiang river water spiked with Hg(2+). In addition, distinguishing aspects of the Fe3O4@nSiO2@mSiO2-T-TRDNA nanocomposites for Hg(2+) detection and removal also include the regeneration using a simple acid treatment and resistance to nuclease digestion. Similar process can be used to functionalize the Fe3O4@nSiO2@mSiO2 nanocomposites with other nucleic acids and small molecules for environmental and biomedical applications.

Intracellular acid-triggered drug delivery system using mesoporous silica nanoparticles capped with T-Hg2+-T base pairs mediated duplex DNA.

This paper proposed a novel intracellular drug delivery system consisting of mesoporous silica nanoparticles (MSN) functionalized on the pore outlets with an acid-labile DNA molecule-gated switch. In this system, a T-Hg2+-T base pair-mediated double-stranded DNA (dsDNA1) was grafted on the MSN surface as a nanoscopic cap. The delivery system was closed at neutral pH but opened at slightly acidic conditions due to the dissociation of T-Hg2+-T structures and the subsequent melting of dsDNA1. As proof-of-concept, doxorubicin (Dox) was loaded into the dsDNA1-modified MSN (MSN-dsDNA1) as a model drug. Controlled-release studies in water showed that no Dox leaked when the cap was closed and that release occurred immediately after acidification. By alternately changing the pH from 5.0 to 7.2, the DNA cap could be switched "on" and "off", thereby regulating the partial release of Dox. Further in vitro studies demonstrated that the Dox-loaded MSN-dsDNA1 (MSN-Dox-dsDNA1) could be endocytosed and accumulated within lysosomes, followed by serving as a carrier for the controlled release of Dox into cell nuclei at the lysosomal pH level inside living cells. The cell viability results showed that the inhibitory concentration (IC50) of MSN-Dox-dsDNA1 was low (≈12.5 µg mL-1), while the IC50 of MSN-dsDNA1 (>100 µg mL-1) exceeded eight times higher than that of MSN-Dox-dsDNA1, indicating that MSN-dsDNA1 was fairly biocompatible and indeed served as a drug-carrier for intracellular controlled release. We believe that further developments of this acid-responsive drug-carrier will provide a promising nanodevice for in vivo delivery of therapeutic agents.

ATP-responsive controlled release system using aptamer-functionalized mesoporous silica nanoparticles.

Adenosine-5'-triphosphate (ATP) is a multifunctional nucleotide, which plays a vital role in many biological processes, including muscle contraction, cells functioning, synthesis and degradation of important cellular compounds, and membrane transport. Thus, the development of ATP-responsive controlled release system for bioorganism application is very significative. Here, an original and facile ATP-responsive controlled release system consisting of mesoporous silica nanoparticles (MSN) functionalized with an aptamer as cap has been designed. In this system, the ATP aptamer was first hybridized with arm single-stranded DNA1 (arm ssDNA1) and arm single-stranded DNA2 (arm ssDNA2) to form the sandwich-type DNA structure and then grafted onto the MSN surface through click chemistry approach, resulting in blockage of pores and inhibition of guest molecules release. In the presence of ATP, the ATP aptamer combined with ATP and got away from the pore, leaving the arm ssDNA1 and ssDNA2 on the surface of MSN. The guest molecules can be released because single-stranded DNA is flexible. The release of the guest molecules from this system then can be triggered by the addition of ATP. As a proof-of-principle, Ru(bipy)(3)(2+) was selected as the guest molecules, and the ATP-responsive loading and release of Ru(bipy)(3)(2+) have been investigated. The results demonstrate that the system had excellent loading efficiency (215.0 µmol g(-1) SiO(2)) and the dye release percentage can reach 83.2% after treatment with 20 mM ATP for 7 h. Moreover, the ATP-responsive behavior shows high selectivity with ATP analogues. However, the leakage of Ru(bipy)(3)(2+) molecule is neglectable if ATP was not added, indicating an excellent capping efficiency. Interestingly, this system can respond not only to the commercial ATP but also to the ATP extracted from living cells. By the way, this system is also relatively stable in mouse serum solution at 37 °C. This proof of concept might promote the application of ATP-responsive devices and can also provide an idea to design various target-responsive systems using other aptamers as cap.

A light-responsive reversible molecule-gated system using thymine-modified mesoporous silica nanoparticles.

In this paper, a reversible light-responsive molecule-gated system based on mesoporous silica nanoparticles (MSN) functionalized with thymine derivatives is designed and demonstrated. The closing/opening protocol and release of the entrapped guest molecules is related by a photodimerization-cleavage cycle of thymine upon different irradiation. In the system, thymine derivatives with hydrophilicity and biocompatibility were grafted on the pore outlets of MSN. The irradiation with 365 nm wavelength UV light to thymine-functionalized MSN led to the formation of cyclobutane dimer in the pore outlet, subsequently resulting in blockage of pores and strongly inhibiting the diffusion of guest molecules from pores. With 240 nm wavelength UV light irradiation, the photocleavage of cyclobutane dimer opened the pore and allowed the release of the entrapped guest molecules. As a proof-of-the-concept, Ru(bipy)(3)(2+) was selected as the guest molecule. Then the light-responsive loading and release of Ru(bipy)(3)(2+) were investigated. The results indicated that the system had an excellent loading amount (53 µmol g(-1) MSN) and controlled release behavior (82% release after irradiation for 24 h), and the light-responsive loading and release procedure exhibited a good reversibility. Besides, the light-responsive system loaded with Ru(bipy)(3)(2+) molecule could also be used as a light-switchable oxygen sensor.