A5100332597- Internet Traffic Analysis and Secure E-voting
- Network Security and Intrusion Detection
- Advanced Malware Detection Techniques
- Advanced Vision and Imaging
- Advanced Image Processing Techniques
- Computer Graphics and Visualization Techniques
- Image Processing Techniques and Applications
- Advanced Neural Network Applications
- Spam and Phishing Detection
- Soft Robotics and Applications
- Robotics and Sensor-Based Localization
- Medical Imaging and Analysis
- Advanced Image and Video Retrieval Techniques
- Blockchain Technology Applications and Security
- Image Enhancement Techniques
- Digital and Cyber Forensics
- Surgical Simulation and Training
- Advanced Steganography and Watermarking Techniques
- Intraocular Surgery and Lenses
- Geotechnical Engineering and Underground Structures
- Retinal and Macular Surgery
- Anomaly Detection Techniques and Applications
- Pelvic and Acetabular Injuries
- Video Coding and Compression Technologies
- Advanced Data Compression Techniques
Institute of Information Engineering
2014-2025
Chinese Academy of Sciences
2014-2025
Institute of Automation
2023-2025
Tongji University
2009-2024
Shandong Institute of Automation
2020-2024
University of Chinese Academy of Sciences
2017-2023
University of Hong Kong
2023
Xiamen University
2023
Changchun Institute of Optics, Fine Mechanics and Physics
2023
Kyushu Institute of Technology
2021-2022
Multidrug resistance (MDR) frequently develops in cancer patients exposed to chemotherapeutic agents and is usually brought about by over-expression of P-glycoprotein (P-gp) which acts as a drug efflux pump reduce the intracellular concentration drug(s). Thus, inhibiting P-gp expression might assist overcoming MDR chemotherapy. MiRNAome profiling using next-generation sequencing identified differentially expressed microRNAs (miRs) between parental K562 cells (K562/ADM) induced adriamycin...
Studies have shown that homeobox D10 (HOXD10) is the target gene of microRNA-10b (miR-10b) and closely associated with inhibition cell migration invasion. Ras homolog family member C (RhoC) has been reported to promote tumor metastasis in various types cancer. The effect miR‑10b on colorectal cancer (CRC) molecular mechanisms remain elusive. present study aimed investigate whether could invasion by targeting HOXD10 CRC exploring association between expression patients. findings revealed...
Depth completion plays a crucial role in autonomous driving, which cameras and LiDARs are two complementary sensors. Recent approaches attempt to exploit spatial geometric constraints hidden enhance image-guided depth completion. However, only low efficiency poor generalization can be achieved. In this paper, we propose BEV@DC, more efficient powerful multi-modal training scheme, boost the performance of practice, proposed BEV@DC model comprehensively takes advantage with rich details...
Although dissolution theory is widely used, in certain circumstance, it seems to be unable explain the hydration of C3S. In this article, more attention paid nucleation products. We find that precipitation C-S-H a nonclassical process. It starts with primary particles and then grows by particle attachment. A sharp increase reaction rate after induction period may come from accelerating growth instead etch pits. The duration relates size floc. Potassium salts influence globule floc mitigate...
In the traditional Natural Language Process (NLP) neighborhood, there is a work that must be discussed called "text classification", which widely used in sentiment analysis, news recommendation, and spam detection. The development of text classification techniques has evolved from manual rules math statistical methods to machine learning deep learning. With advancements self-supervised multimodal learning, becoming more intelligent efficient. This study focuses on exploring application...
Compositional generalization is the capability of a model to understand novel compositions composed seen concepts. There are multiple levels including phrase-phrase level, phrase-word and word-word level. Existing methods achieve promising compositional generalization, but consistency across remains unexplored. The refers that should generalize level composition, phrase-word/word-word can be derived from it simultaneously. In this paper, we propose meta-learning based framework, for...
In this paper, we formulate the deep residual network (ResNet) as a control problem of transport equation. ResNet, equation is solved along characteristics. Based on observation, neural closely related to PDEs manifold. We propose several models based and Hamilton-Jacobi The discretization these point cloud also discussed.
In order to protect user privacy or guarantee free access the Internet, network covert channel has become a hot research topic. It refers an information in which messages are covertly transmitted under environment. recent years, many new construction schemes of channels proposed. But at same time, also received attention censors, leading attacks. The Many users exploit Internet. Previous based on steganography, can be divided into CTCs and CSCs. there some constructed by changing...
Inverse rendering of complex material such as glossy, metal and mirror is a long-standing ill-posed problem in this area, which has not been well solved. Previous approaches cannot tackle them due to simplified BRDF unsuitable illumination representations. In paper, we present PhyIR, neural inverse method with more completed SVBRDF representation physics-based in-network layer, can handle incorporate physical constraints by re-rendering realistic detailed specular reflectance. Our framework...
As the tool for launching cyber attacks, ever-increasing malware variants pose a significant threat to interconnected network community. The detection methods based on conventional machine learning techniques require lots of samples training. However, in real-world scenarios, such as early stage novel attacks appearance, only small number malicious can be obtained. Applying data-intensive traditional above scenarios will cause serious overfitting problems. Therefore, there is need few-shot...
We present a efficient multi-view inverse rendering method for large-scale real-world indoor scenes that reconstructs global illumination and physically-reasonable SVBRDFs. Unlike previous representations, where the of large is simplified as multiple environment maps, we propose compact representation called Texture-based Lighting (TBL). It consists 3D mesh HDR textures, efficiently models direct infinite-bounce indirect lighting entire scene. Based on TBL, further hybrid with precomputed...