A5101858043- Face recognition and analysis
- Sperm and Testicular Function
- Forensic and Genetic Research
- Advanced Optimization Algorithms Research
- Osteoarthritis Treatment and Mechanisms
- Artificial Intelligence in Healthcare
- High Altitude and Hypoxia
- Anomaly Detection Techniques and Applications
- Face and Expression Recognition
- Digital Media Forensic Detection
- COVID-19 diagnosis using AI
- Robotic Path Planning Algorithms
- Insect symbiosis and bacterial influences
- Optimization and Mathematical Programming
- interferon and immune responses
- Multi-Criteria Decision Making
- Ferroptosis and cancer prognosis
- Metaheuristic Optimization Algorithms Research
- Generative Adversarial Networks and Image Synthesis
- AI in cancer detection
- Advanced biosensing and bioanalysis techniques
- Digital Imaging for Blood Diseases
- Cancer, Hypoxia, and Metabolism
- Imbalanced Data Classification Techniques
- Guidance and Control Systems
Kunming Medical University
2025
Tianjin Medical University
2011-2025
Hong Kong Baptist University
2025
University of Notre Dame
2024
Chinese Academy of Agricultural Sciences
2024
Institute of Plant Protection
2024
Southern University of Science and Technology
2023-2024
Zhongnan Hospital of Wuhan University
2024
Wuhan University
2024
Beijing Normal University - Hong Kong Baptist University United International College
2023
A variety of coordinated host-cell responses are activated as defense mechanisms against pore-forming toxins (PFTs). Bacillus thuringiensis (Bt) is a worldwide used biopesticide whose efficacy and precise application methods limits its use to replace synthetic pesticides in agricultural settings. Here, we analyzed the intestinal two lepidopteran insect pests after intoxication with sublethal dose Bt PFTs find out potential functional genes. We show that larval epithelium was initially...
Abstract Efficient intracellular protein delivery is of great importance for the development protein‐based therapy and modern biotechnologies. However, hydrophilic macromolecular nature proteins greatly hinders their ability to cross cell membranes. Herein, a calixarene modification strategy drugs developed. The decoration sulfonate azocalix[4]arene (SAC4A) on results in nano‐multivalent effect between Protein‐S amino acids surface, leading efficient via clathrin‐mediated endocytic pathway....
Distinguishing osteosarcoma from bone callus remains a clinical challenge due to their morphological similarities. This study proposes J-CAN, multimodal deep learning framework integrating hyperspectral imaging (HSI) and H&E-stained pathology for rapid accurate classification. The HSI system captures 176 spectral bands (400-1000 nm), providing molecular-level insights. MobileNetV2 extracts spatial features, while 1D-CNN processes signatures. A self-attention mechanism enhances feature...
Sepsis-associated encephalopathy (SAE) is a serious neurological complication accompanied by acute and long-term cognitive dysfunction. Ferroptosis newly discovered type of cell death that produced iron-dependent lipid peroxidation. As key transcriptional coactivator in the Hippo signalling pathway, Yes-associated protein 1 (YAP1) could target ferroptosis-related genes. This study was aimed to determine whether Yap1 protects against SAE inhibits ferroptosis via maintaining mitochondrial...
To study the systemic production of inflammatory factors and activation transcription factor nuclear kappa B (NF-κB) in response to different levels intermittent hypoxia time.A total 160 male Wistar rats were divided randomly into five groups. The first three groups exposed 5%, 7.5% 10% (referred as IH-1, IH-2, IH-3 respectively), fourth group subjected sustained (abbreviated SH), control normal oxygen (designated SC). At second, fourth, sixth, eighth week, eight each sacrificed collect...
In recent years, deepfake videos have been abused to create fake news, which threaten the integrity of digital videos. Although existing detection methods leveraged cumbersome neural networks achieve promising performance, they cannot be deployed in resource-constrained scenarios. To overcome this limitation, we propose a novel model compression framework based on joint distillation for detection, includes pre-training stage and knowledge transfer stage. stage, teacher network is trained...
Loop Closure Detection is of great significance in the field intelligent driving systems, as it reduces cumulative error estimated position system and assists generating a consistent global map. Existing methods differ frame representation corresponding frame-matching strategy. Traditionally, local feature points descriptors are studied extensively while recently semantic information extracted from deep learning considered superior terms promoting high-level understanding surrounding...
The context of the COVID-19 pandemic has made an increasing number people accustomed to wearing masks avoid getting infected. This phenomenon leads urgent demand for face recognition with masks, which poses serious challenge accuracy recognition. paper seeks produce a model called AFF-Net based on attention mechanism and feature fusion self-attention address masked by designing four compact modules achieve optimization network structure solve problem unstable rate low generalization ability....
Cutting planes are crucial in solving mixed integer linear programs (MILP) as they facilitate bound improvements on the optimal solution. Modern MILP solvers rely a variety of separators to generate diverse set cutting by invoking frequently during process. This work identifies that can be drastically accelerated appropriately selecting activate. As combinatorial separator selection space imposes challenges for machine learning, we learn separate proposing novel data-driven strategy restrict...
Male infertility accounts for about one-third of global cases. Manual assessment sperm abnormalities through head morphology analysis encounters issues observer variability and diagnostic discrepancies among experts. Its alternative, Computer-Assisted Semen Analysis (CASA), suffers from low-quality images, small datasets, noisy class labels. We propose a new approach classification, called SHMC-Net, which uses segmentation masks heads to guide the classification images. SHMC-Net generates...
Mixed Integer Linear Programming (MILP) is essential for modeling complex decision-making problems but faces challenges in computational tractability and requires expert formulation. Current deep learning approaches MILP focus on specific problem classes do not generalize to unseen classes. To address this shortcoming, we take a foundation model training approach, where train single diverse set of across As existing datasets lack diversity volume, introduce MILP-Evolve, novel LLM-based...
The pursuit-evasion problem is widespread in nature, engineering and societal applications. It commonly observed nature that a predator runs faster than its prey but it has less agile maneuverability. Over millions of years evolution, animals have developed effective efficient pursuit evasion strategies. In this paper, we provide dynamic framework for unicycle systems from nature-inspired perspective. Firstly, the with one pursuer evader, propose an Alert-Turn control strategy which consists...