Article
Geography, Physical
Shiqi Tian, Xicheng Tan, Ailong Ma, Zhuo Zheng, Liangpei Zhang, Yanfei Zhong
Summary: Remote sensing imagery combined with deep learning enables change detection in land cover and land use. However, the high similarity of ground object features in temporal high-resolution remote sensing imagery makes change identification difficult. This paper proposes TCRPN, a method that highlights change regions using saliency from single-temporal images to enhance spatio-temporal features, and improves change representation distinguishability through shortcuts.
ISPRS JOURNAL OF PHOTOGRAMMETRY AND REMOTE SENSING
(2023)
Article
Agriculture, Multidisciplinary
Rujing Wang, Lin Jiao, Chengjun Xie, Peng Chen, Jianming Du, Rui Li
Summary: This study proposes an effective method for small pest recognition and detection by introducing attention mechanism and redesigning the region proposal network. Experimental results demonstrate that the proposed method outperforms other existing pest detection methods on the AgriPest21 dataset.
COMPUTERS AND ELECTRONICS IN AGRICULTURE
(2021)
Article
Agriculture, Multidisciplinary
Xue Zhao, Kaiyu Li, Yunxia Li, Juncheng Ma, Lingxian Zhang
Summary: This study developed a new vegetable disease identification model, DTL-SE-ResNet50, and compared it with other models. The results showed that the new model had high identification precision and fast identification speed.
COMPUTERS AND ELECTRONICS IN AGRICULTURE
(2022)
Article
Agriculture, Multidisciplinary
Dandan Wang, Dongjian He
Summary: This study developed a precise apple instance segmentation method based on an improved Mask RCNN, which achieved accurate apple segmentation under various conditions and demonstrated near real-time performance. The method outperformed other comparison methods and laid the foundation for accurate fruit detection and long-term automatic growth monitoring.
COMPUTERS AND ELECTRONICS IN AGRICULTURE
(2022)
Article
Computer Science, Information Systems
Wei Ma, Helong Yu, Wenbo Fang, Fachun Guan, Dianrong Ma, Yonggang Guo, Zhengchao Zhang, Chao Wang
Summary: Timely detection, identification, and assessment of crop diseases are crucial for disease prevention and control. This study proposes a neural network-based method that utilizes an improved Rouse spatial pyramid pooling strategy to achieve accurate and high-resolution crop disease detection. The results indicate that the proposed method outperforms conventional methods in terms of accuracy and noise suppression.
Article
Ecology
Enlin Li, Liwei Wang, Qiuju Xie, Rui Gao, Zhongbin Su, Yonggang Li
Summary: Maize diseases have a significant impact on crop yield, but identifying them on a large scale has been challenging due to limited human experience and traditional image recognition technology. However, deep learning-based methods offer promise for automatic disease identification. In this study, a deep learning-based method using the MDCDenseNet model was proposed for maize disease identification. The model outperformed other models with an accuracy of 98.84% when tested on field-collected datasets with complex backgrounds. This approach provides a viable solution for identifying maize leaf diseases with small sample sizes and complex backgrounds.
ECOLOGICAL INFORMATICS
(2023)
Article
Computer Science, Artificial Intelligence
Jinsong Su, Jialong Tang, Hui Jiang, Ziyao Lu, Yubin Ge, Linfeng Song, Deyi Xiong, Le Sun, Jiebo Luo
Summary: In this study, a progressive self-supervised attention learning approach is proposed to enhance the performance of aspect-based sentiment analysis (ABSA) models by continuously learning useful attention supervision information through iterative processes. Experimental results demonstrate that this method can effectively improve the attention mechanism of ABSA models and enhance their performance.
ARTIFICIAL INTELLIGENCE
(2021)
Article
Computer Science, Information Systems
Junquan Meng, Feng Kang, Yaxiong Wang, Siyuan Tong, Chenxi Zhang, Chongchong Chen
Summary: This study proposes an improved YOLOv7 algorithm by using Depth Separable Convolution (DS Conv) blocks, Convolutional Block Attention Modules (CBAM), and Coordinate Attention (CA) modules to solve the problem of difficult identification of tea buds due to their similar color with the background in complex scenes. The method improves the mean Average Precision (mAP) by 1.28% and mean Recall (mR) rate by 2.92%, achieving final mAP and mR of 96.70% and 93.88% respectively. The improved model also meets real-time detection requirements with a frame rate of 30.62 FPS. The results show that the improved YOLOv7 algorithm has higher detection accuracy for tea buds compared to other target detection algorithms, and it performs well under different light conditions, providing valuable insights for intelligent tea picking.
Article
Environmental Sciences
Han Liang, Suyoung Seo
Summary: Semantic segmentation of remote sensing images is crucial in urban planning and development. This paper proposes a lightweight progressive attention semantic segmentation network that reduces computational costs without sacrificing accuracy, addressing the challenges of semantic segmentation of remote sensing images.
Article
Agronomy
Wei Sun, Chunshan Wang, Jingqiu Gu, Xiang Sun, Jiuxi Li, Fangfang Liang
Summary: The plant disease recognition model based on deep learning has good potential, but low transparency and poor interpretability limit its deployment and application in field scenarios. To address these limitations, we propose Veg DenseCap, a model that takes vegetable leaf images, locates abnormal parts, and generates description statements for disease features in natural language. It achieves better performance compared to the classical FCLN model.
Article
Environmental Sciences
Jinsheng Xiao, Haowen Guo, Yuntao Yao, Shuhao Zhang, Jian Zhou, Zhijun Jiang
Summary: In this paper, a new method for image object detection is proposed, which improves the detection of small objects in complex backgrounds through multi-feature selection and fusion. The use of an attention mechanism-based multi-feature selection module reduces the interference of useless information and improves detection accuracy.
Article
Computer Science, Artificial Intelligence
Jia Zhang, Wei Li, Zhixin Li
Summary: Unsupervised domain adaptive semantic segmentation utilizes knowledge learned from labeled source domain dataset to guide segmentation in the target domain. Pseudo labels are generated using a self-supervised learning method to align corresponding pixels with the source domain based on segmentation loss. A channel and spatial parallel attention module is employed to extract rich spatial and channel information from the feature map, and focal loss is introduced to address class imbalance in the dataset.
IMAGE AND VISION COMPUTING
(2022)
Article
Computer Science, Artificial Intelligence
He Yao, Yongjun Zhang, Huachun Jian, Li Zhang, Ruzhong Cheng
Summary: This study addresses the issue of suboptimal single-spectral nighttime pedestrian detection performance by incorporating background information into the channel attention mechanism. By adopting a contrast learning approach, the proposed Fore-Background Contrast Attention (FBCA) method significantly outperforms existing methods in nighttime pedestrian detection and achieves state-of-the-art results on multiple datasets.
KNOWLEDGE-BASED SYSTEMS
(2023)
Article
Computer Science, Information Systems
Lujuan Deng, Boyi Liu, Zuhe Li, Jiangtao Ma, Hanbing Li
Summary: Multimodal sentiment analysis aims to understand people's attitudes and opinions from different data forms. This article proposes a new model based on a recurrent neural network with a complex attention mechanism to fully utilize context information and the correlation between modalities. The model effectively captures semantic information and contextual relationship and fuses different pieces of modal information.
Article
Chemistry, Multidisciplinary
Wenbo Zhu, Quan Wang, Lufeng Luo, Yunzhi Zhang, Qinghua Lu, Wei-Chang Yeh, Jiancheng Liang
Summary: This paper proposes a new attention mechanism (CPAM) to address the issue of regional bias in tile block defect detection. By dividing feature information into patches, CPAM can successfully distinguish different regional features and linearly connect these patches in two spatial directions, thereby improving the performance of the model.
APPLIED SCIENCES-BASEL
(2022)
Article
Agriculture, Multidisciplinary
Chunshan Wang, Ji Zhou, Chunjiang Zhao, Jiuxi Li, Guifa Teng, Huarui Wu
Summary: This paper proposes a small-sample recognition model of vegetable diseases in complex backgrounds based on image text collaborative representation learning (ITC-Net), which achieves good results. The model combines disease image modal information with disease text modal information, effectively utilizing the correlation and complementarity between the two types of disease information.
COMPUTERS AND ELECTRONICS IN AGRICULTURE
(2021)
Article
Environmental Sciences
Chunshan Wang, Qian Wang, Huarui Wu, Chunjiang Zhao, Guifa Teng, Jiuxi Li
Summary: The proposed method in this paper enhanced the accuracy of small target detection in poppy inspection by adding a larger detection box and optimizing model parameters. Test results showed that the new model outperformed traditional models and improved work efficiency in handling large-scale image data.
Article
Agriculture, Multidisciplinary
Chunshan Wang, Pengfei Du, Huarui Wu, Jiuxi Li, Chunjiang Zhao, Huaji Zhu
Summary: The study proposed a two-stage model that combines DeepLabV3+ and U-Net for cucumber leaf disease severity classification in complex backgrounds. Experimental results showed that the model was able to segment leaves and disease spots step-by-step to complete the disease severity classification with high accuracy.
COMPUTERS AND ELECTRONICS IN AGRICULTURE
(2021)
Article
Agriculture, Multidisciplinary
Ji Zhou, Jiuxi Li, Chunshan Wang, Huarui Wu, Chunjiang Zhao, Guifa Teng
Summary: This paper focused on the recognition of common invasive diseases in tomatoes and cucumbers. By utilizing multimodal data and domain knowledge, a disease identification model based on image-text collaborative representation and knowledge assistance was constructed, achieving high accuracy and sensitivity.
COMPUTERS AND ELECTRONICS IN AGRICULTURE
(2021)
Article
Geology
Yi-Wei Peng, Hao Zou, Leon Bagas, Yu-Fan Shen, Zhi-Ping Shu, Jing Su, Qing-Dong Liang, Chun-Shan Wang, Yao-Hua Hu, Heng Zhang
Summary: This study used geological and isotope analysis to determine the age of mineralization, source of ore-forming components, and the relationship between mineralization and magmatic activity in the Arqiale Pb-Zn-Cu deposit.
ORE GEOLOGY REVIEWS
(2022)
Article
Environmental Sciences
Qian Wang, Chunshan Wang, Huarui Wu, Chunjiang Zhao, Guifa Teng, Yajie Yu, Huaji Zhu
Summary: This study proposed a two-stage method for detecting illegal opium poppy cultivation sites, effectively reducing the workload for manual screening.
Article
Plant Sciences
Xuguang Feng, Chunjiang Zhao, Chunshan Wang, Huarui Wu, Yisheng Miao, Jingjian Zhang
Summary: In this paper, an end-to-end disease identification model combining a disease-spot region detector and a disease classifier was proposed for automatic identification of vegetable diseases in field environments. By introducing bidirectional cross-modal feature fusion, the model achieved optimal results on a small dataset.
FRONTIERS IN PLANT SCIENCE
(2022)
Article
Environmental Sciences
Chunshan Wang, Wei Sun, Huarui Wu, Chunjiang Zhao, Guifa Teng, Yingru Yang, Pengfei Du
Summary: This study proposes a low-altitude remote sensing method based on a modified YOLOv5s-ViT model for detecting rural living environments. By modifying the BottleNeck structure, embedding the SimAM attention mechanism module, and incorporating the Vision Transformer component, the model's feature capture capability and perception ability are improved. Experimental results show that the modified model achieves improvements in Precision, Recall, and mAP compared to the original model, while reducing the number of parameters and computation volume. This study provides new ideas for enhancing the digital capability of governance in rural living environments.
Article
Agronomy
Chunshan Wang, Shedong Sun, Chunjiang Zhao, Zhenchuan Mao, Huarui Wu, Guifa Teng
Summary: This study proposes a cucumber root-knot nematode detection model based on the modified YOLOv5s model (YOLOv5-CMS), which enhances the model's ability to extract key features, optimizes the cluster algorithm, and improves the detection precision. Experimental results show that the YOLOv5s-CMS model achieves better performance compared to the original model, providing more intuitive and accurate data sources for breeding cucumber varieties resistant to root-knot nematode.
Article
Agronomy
Wenhao Zhang, Chunshan Wang, Huarui Wu, Chunjiang Zhao, Guifa Teng, Sufang Huang, Zhen Liu
Summary: An entity-relation-extraction model for crop diseases (BBCPF) was proposed in this paper, utilizing the advantages of knowledge graphs. Through the use of BERT, BiLSTM, and CRF models, the precision and recall values were optimized for entity and relation extraction, providing an effective method for knowledge graph construction in the Chinese crop disease domain.
Article
Chemistry, Analytical
Bin Wang, Yan Zhang, Chunshan Wang, Guifa Teng
Summary: The aim of this study is to establish a method for real-time calculating droplet deposition distribution of a six-rotor plant protection unmanned aerial vehicle (UAV). Numerical simulation and statistical calculations were used to analyze the airflow field and droplet deposition rates under different parameter combinations. The study found that relative airflow and crosswind affected the droplet distribution, and a prediction method was established based on these distributions.
Article
Plant Sciences
Chunshan Wang, Ji Zhou, Yan Zhang, Huarui Wu, Chunjiang Zhao, Guifa Teng, Jiuxi Li
Summary: The study addressed the issue of weak robustness in disease image recognition models based on deep learning by proposing a feature decomposition and recombination method, and applying graph convolutional neural network for feature learning to build a vegetable disease recognition model based on the fusion of images and graph structure text.
FRONTIERS IN PLANT SCIENCE
(2022)
Article
Computer Science, Information Systems
Chunshan Wang, Siqi He, Huarui Wu, Guifa Teng, Chunjiang Zhao, Jiuxi Li
Article
Computer Science, Information Systems
Chunshan Wang, Ji Zhou, Huarui Wu, Jiuxi Li, Zhao Chunjiang, Rong Liu