A5100653012- Video Surveillance and Tracking Methods
- Advanced Neural Network Applications
- Smart Agriculture and AI
- Human Pose and Action Recognition
- Network Security and Intrusion Detection
- Educational Innovations and Challenges
- Advanced Image and Video Retrieval Techniques
- Infrared Target Detection Methodologies
- Immunotherapy and Immune Responses
- Computational Geometry and Mesh Generation
- Graph Theory and Algorithms
- Underwater Acoustics Research
- Human Motion and Animation
- Spam and Phishing Detection
- Machine Learning and Data Classification
- Industrial Vision Systems and Defect Detection
- Earthquake Detection and Analysis
- Neuroscience, Education and Cognitive Function
- Biometric Identification and Security
- Spectroscopy and Chemometric Analyses
- Remote-Sensing Image Classification
- Remote Sensing in Agriculture
- Speech and Audio Processing
- Handwritten Text Recognition Techniques
- Regional Economic Development and Innovation
Sino Biological (China)
2025
Shandong Agricultural University
2022-2023
Shenzhen University
2022
Sichuan Agricultural University
2022
Anhui University
2021
Southwest Jiaotong University
2021
Chinese Academy of Sciences
2021
Tianjin Renai College
2021
Huazhong University of Science and Technology
2017-2020
Nanjing University of Science and Technology
2020
Pseudorabies (PR) caused by the re-emerging of pseudorabies virus (PRV) variant has outbroken among PRV vaccine immunized swine in many pig farms, which serious social and economic consequences since end 2011. The UL49.5 protein can inactivate transporter associated with antigen processing (TAP), thereby downregulating cell surface expression leukocyte class I (SLA-I) to evade host immune surveillance. In this study, based on ΔgE/TK strain, ΔgE/TK/UL49.5 triple gene deletion strain was...
The prevention and management of crop diseases play an important role in agricultural production, but there are many types complex causes, their identification add difficulties to the process. traditional methods identifying mostly rely on human visual manual inspection, which requires a certain amount expert knowledge experience. There shortcomings such as strong subjectivity low accuracy. This paper takes common kiwifruit research object. Based deep learning computer vision models, given...
Abstract A cascade reaction involving a base-promoted nucleophilic substitution between diketones and α-bromoacetophenone derivatives the subsequent selective condensation–cyclization was developed for synthesis of 1,2-diaroyl benzofurans. 1,2-Diaroyl benzofurans with different functional groups structures exhibit reversible photochromic behaviors in solution, solid state, thin films diverse colors, demonstrating potential application field optical materials.
Deep neural networks have achieved promising performance for hyperspectral image (HSI) classification. However, due to the limitation of available labeled samples, traditional deeper and wider usually cause overfitting problem lose detailed information. To solve this problem, a brain-like structure, namely spatial attention-driven recurrent feedback convolutional network (SARFNN), is proposed by utilizing attention mechanism structures, from which two deep models are further developed HSI...
Path planning of mobile robots is an important research content in robot design. In view the characteristics classic ant colony algorithm, such as long distance, low efficiency and easy to fail into dead zone, a new pheromone updating algorithm global/local distribution models are proposed. The grid method used construct two-dimensional plane space with obstacles simulation experiment carried out. results show that under improved path high distance short, which verifies validity model.
Recently, various object detection frameworks have been applied to text tasks and achieved good performance in the final detection. With further expansion of application scenarios, research value topics has gradually increased. Text natural scenes is more challenging for horizontal based on a quadrilateral box curved any shape. Most networks effect balancing target samples detection, but it deal with small targets solve extremely unbalanced data. We continued use PSENet such problems this...
Deep semi-supervised learning is a hot research topic in recent years. The main challenges are the small sample and make full use of unlabeled data. Pseudo-labeling simple yet effective approach learning. However, how to obtain high quality pseudo-labeled data key issue. When with low confidence level used train deep model will enlarge or spread classification errors. To overcome this issue, we propose confidence-based pseudo-labeling method choose pseudo-labels novel uncertainty model....
On the basis of building an on-campus and off-campus high-level collaborative innovation platform, Hunan University Humanities, Science Technology organized dual-tutor team, explored united cultivation applied postgraduates in agriculture with enterprise public institution, which enhanced adaptability services for local social economic development our college, promoted service ability level.Thus, there was a marked improvement quality postgraduates.
Based on the elaboration of basic connotation for collaborative innovation local colleges and universit ies, paper analyzes dilemmas that universities faced in environment, indicates is best effective approach promotion scientific technological service ability universities, meanwhile, it also suggests strengthening consciousness innovation, building p latform exploring mechanism so as to provide reference universities.
Aiming at the problem of low intelligence in automatic navigation cuttage and film covering multi-functional machine for tunnels, this study proposed a line extraction method based on improved YOLOv5s model, which can achieve accurate lines two planting methods seedling transplanting direct seeding. Firstly, we pre-processed acquired images using inverse perspective transformation. Next, Coordinate Attention Ghost modules were applied to improve architecture, increasing detection accuracy...
By introducing the nonholonomic constraints, natural gradient algorithm is effective to overcome shortcomings of traditional algorithm. Namely, when source signal amplitude changes rapidly over time or equal zero in a certain period time, it can still work well. In addition, selecting different estimate function stage get balance between convergence speed and steady-state performance. Thus, this paper combines with optimal selective obtain new Furthermore, variable step-size which based on...
目的 生物质炭显著影响土壤氧化亚氮 (N2O) 排放，但关于其相关微生物机理的研究相对匮乏，尤其是生物质炭对酸性菜地土壤N2O排放的微生物作用机理。本文通过研究氮肥配施生物质炭对酸性菜地土壤N2O排放以及硝化和反硝化过程相关功能基因丰度的影响，探讨酸性菜地土壤N2O排放与功能基因丰度的关系，阐释生物质炭对酸性菜地土壤试验N2O排放的微生物作用机理。 方法 在田间一次性施入生物质炭 40 t/hm2，试验连续进行了3年，共9茬蔬菜。设置4个处理：对照 (CK)、氮肥 (N)、生物质炭 (Bc) 和氮肥 + 生物质炭 (N Bc)。在施用后第三年，采集土壤样品进行室内培养，应用荧光定量PCR技术检测硝化过程氨氧化古菌 (AOA)、氨氧化细菌 (AOB) 功能基因amoA和反硝化过程亚硝酸还原酶基因 (nirK、nirS) 以及N2O还原酶基因 (nosZ) 等相关功能基因丰度，同时监测土壤pH值、无机氮 (铵态氮、硝态氮) 含量及N2O排放。 结果 与CK相比，生物质炭 处理的土壤有机碳 (SOC) 提高了27.1%，总氮...