Xuanjiang Yang’s research while affiliated with Hefei Institutes of Physical Science and other places

... Huang et al. [22] introduced a method that combines the Lazy Eye algorithm with distance transform to improve the D* Lite algorithm, significantly enhancing path safety and smoothness. Xu [23] proposed a fusion algorithm with layered planning, using D* Lite for global planning and enhanced neural network algorithms for local planning, greatly optimizing path planning time. Zhang et al. [24] divided an environmental map into cells, set core grids within the cells, and constructed search linked lists sequentially to guide the correct search direction, improving the path planning efficiency of the D* Lite algorithm. ...

Reference:

Improved D* Lite Algorithm for Ship Route Planning

... In the context of pesticide detection, the technique exploits the unique spectral signatures associated with specific functional groups in pesticide molecules. By illuminating a sample with near-infrared light and analyzing the resulting absorption or reflection spectra, NIR allows for the identification and quantification of pesticides in a variety of matrices, including crops [155], soil [156], and water [157]. ...

Reference:

A Review of Emerging Techniques for Pyrethroid Residue Detection in Agricultural Commodities

... The feature pyramid network (FPN) [17] utilized in Faster R-CNN and Mask R-CNN [18] is shown in Figure 1a. It uses the features of the five stages of the ResNet convolution groups C2-C6, among which C6 is obtained from a MaxPooling operation by directly applying 1 × 1/2 on C5. ...

Reference:

Two Novel Models for Traffic Sign Detection Based on YOLOv5s

... Lou et al. [23] proposed a one-stage detection model called YOLOv5 algorithm for detecting cucumber leaf diseases based on deep learning for NVIDIA Jetson Xavier NX. Their model obtained 84.6 accuracy result mAP on the constructed cucumber leaf image dataset after labeling. ...

Reference:

Lightweight improved yolov5 model for cucumber leaf disease and pest detection based on deep learning

... Image segmentation for plant disease enables localization of disease-infected organs Lakshmi and Nickolas, 2016;Akanksha et al., 2021;Chouhan et al., 2021). To this end, Zhong et al. (2021) provided a multi-modal plant disease image dataset (PDID) with classification, detection, and segmentation labels. The authors also proposed a triple stream segmentation network for robust plant disease segmentation. ...

Reference:

Plant Organellar Omics

... The Boundary-Aware Network (BANet) [31] proposed a two-stream framework to achieve semantic segmentation which is not completely separated based on BiSeNet's feature fusion module. The Dual Stream Segmentation Network (DSSNet) [32] introduced its attention module and pyramid pooling module based on BiSeNet. Inspired by BiSeNet, TB-Net [33] proposed a three-stream boundary aware network which changes the context path with a context-aware attention module and adds Boundary Stream to enhance the segmentation performance, particularly for thin and small objects. ...

Reference:

DuFNet: Dual Flow Network of Real-Time Semantic Segmentation for Unmanned Driving Application of Internet of Things

... Considering the problems of the presence of antibiotics in the environment, we need to remove them from the environment by different methods (Polianciuc et al. 2020). The methods for removal of antibiotics such as AMX and AMC include the effluent treatment process, coagulation/sediment, adsorption (Fe 3 O 4 /activated carbon/chitosan adsorbents), chlorination, ozonation, advanced oxidation process (AOPs), biomass ash, and microbial bioremediation (Conde-Cid et al. 2019Jiang et al. 2023;Mert et al. 2018;Santás-Miguel et al. 2020;Yang et al. 2020b). ...

Reference:

Removal of amoxicillin and co-amoxiclav by newly isolated Stenotrophomonas maltophilia DF1

... Rodriguez Alvarez et al. [36,37] employed convolutional neural networks (CNNs) for BCS estimation, with one study incorporating ensemble modeling for improved accuracy [37]. Huang et al. [8] utilized a single-shot multibox detector (SSD) for BCS detection, while Li et al. [38] combined YOLOv2 with CNN for BCS prediction. Çevik [39] further developed a real-time classification system based on pre-trained deep learning architectures and a mobile application for dairy farmers. ...

Reference:

Automated Classification System Based on YOLO Architecture for Body Condition Score in Dairy Cows

... In terms of 2D imaging, Huang et al. (2019) [9] introduced an improved SSD algorithm, achieving model lightweighting by replacing network connections and transforming convolution methods, ultimately keeping the model size at 23.1 MB while enhancing performance. Feng et al. (2024) [10] proposed a lightweight model that integrated attention modules to enhance focus on key features and replaced activation functions to reduce computational resource consumption, achieving a lightweight design. ...

Reference:

Efficient Cow Body Condition Scoring Using BCS-YOLO: A Lightweight, Knowledge Distillation-Based Method

... Using the SVM model, an accuracy of 73.33% was obtained. Wang et al. [16] applied the transfer learning technique to train a machine learning model to detect plant disease and weed identification. Using transfer learning and the combination of multiple deep CNN, authors were able to achieve promising results. ...

Reference:

Genetic Algorithm–Aided Deep Feature Selection for Improved Rice Disease Classification