A5100759164- Face and Expression Recognition
- Video Surveillance and Tracking Methods
- Image Retrieval and Classification Techniques
- Advanced Image and Video Retrieval Techniques
- Human Pose and Action Recognition
- Advanced Measurement and Detection Methods
- Anomaly Detection Techniques and Applications
- Speech and Audio Processing
- Advanced Data Compression Techniques
- Advanced Decision-Making Techniques
- Recommender Systems and Techniques
- Face recognition and analysis
- Human Mobility and Location-Based Analysis
- Fire Detection and Safety Systems
- Chaos-based Image/Signal Encryption
- Evaluation Methods in Various Fields
- Image and Video Stabilization
- Traffic Prediction and Management Techniques
- Wireless Communication Networks Research
- Single-cell and spatial transcriptomics
- Remote Sensing and Land Use
- Advanced Statistical Methods and Models
- Industrial Vision Systems and Defect Detection
- Remote Sensing and LiDAR Applications
- Optical measurement and interference techniques
Beijing Institute of Technology
2013-2024
Huaqiao University
2012-2024
Beijing University of Posts and Telecommunications
2023-2024
Switch
2023
Institute of Basic Medical Sciences of the Chinese Academy of Medical Sciences
2023
Anhui Normal University
2021
BaiCheng Normal University
2019
The University of Tokyo
2018
Bournemouth University
2018
Xidian University
2017
Despite the great success of Siamese-based trackers, their performance under complicated scenarios is still not satisfying, especially when there are distractors. To this end, we propose a novel Siamese relation network, which introduces two efficient modules, i.e. Relation Detector (RD) and Refinement Module (RM). RD performs in meta-learning way to obtain learning ability filter distractors from background while RM aims effectively integrate proposed into framework generate accurate...
The acceleration of urbanization and the increasing demand for precise city planning have made extraction buildings roads from remote sensing images crucial. Deep learning-based methods propelled progress object technology, but there are still challenges such as missing incomplete small objects occlusions. To address this issue, we propose a dual-path network based on CNN Transformer, combining local global features to fully extract semantic information objects. further enhance...
Abstract The sparsity problem remains a significant bottleneck for recommendation systems. In recent years, deep matrix factorization has shown promising results in mitigating this issue. Furthermore, many works have improved the prediction accuracy of by incorporating user’s and/or items’ auxiliary information. However, there are still two remaining drawbacks that need to be addressed. First, initialization latent feature representations substantial impact on performance factorization, and...
Due to the complexity of microservice architecture, it is difficult accomplish efficient anomaly detection and localization tasks achieve target high system reliability. For rapid failure recovery user satisfaction, significant detect locate anomalies fast accurately in systems. In this paper, we propose an model based on Transformer, named TADL (Transformer-based Anomaly Detector Locator), which models temporal features dynamically captures container relationships using Transformer with...
This paper presents a novel lossless compression technique of the context-based adaptive arithmetic coding which can be used to further compress quantized parameters in audio codec. The key feature new is combination context model time domain and frequency called time-frequency model. It for such as modified discrete cosine transform (MDCT) coefficients band gains ITU-T G.719 With proposed coding, high degree adaptation redundancy reduction achieved. In addition, an efficient variable rate...
In the recent years, manifold learning methods have been widely used in data classification to tackle curse of dimensionality problem, since they can discover potential intrinsic low-dimensional structures high-dimensional data. Given partially labeled data, semi-supervised algorithms are proposed predict labels unlabeled points, taking into account label information. However, these not robust against noisy especially when contain noise. this paper, we propose a framework for (RSSML) address...
N6-methyladenosine (m6A)-mediated epitranscriptomic regulation is critical for various physiological processes. Genetic studies demonstrate that proper m6A-methylation required mouse brain development and function. Revealing landscapes of in the cerebral cortex at different developmental stages will help to understand biological meaning regulation. Here, we depict temporal-specific status embryonic postnatal cortices using methylated RNA immunoprecipitation (MeRIP) sequencing. We identified...
Nonnegative matrix factorization (NMF), a well-known technique to find parts-based representations of nonnegative data, has been widely studied. In reality, ordinal relations often exist among such as data i is more related j than q. Such relative order naturally available, and importantly, it truly reflects the latent structure. Preserving enables us structured that are faithful order, so learned become discriminative. However, current NMFs pay no attention this. this paper, we make first...
For 3D point cloud data scanned by a laser scanner, noise has become an obstacle for reconstruction. Though there are many papers about de-nosing, few based on the scanner. Now we propose some methods processing base scanning data. In this paper frequently use way to process generated machine. We will deal with clustering and sampling steps finally achieve satisfying de-noising result.
Trisomy 18, commonly known as Edwards syndrome, is the second most common autosomal trisomy among live born neonates. Multiple tissues including cardiac, abdominal, and nervous systems are affected by an extra chromosome 18. To delineate complexity of anomalies we analyzed cultured amniotic fluid cells from two euploid three 18 samples using single-cell transcriptomics. We identified 6 cell groups, which function in development major such kidney, vasculature smooth muscle, display...
Abstract The accurate prediction of individual travel mode in the city is necessary for development urban traffic intelligence. Through deep analysis on distribution modes, policy makers can make proper policies to improve conditions. With sensing techniques and telecommunication technologies, it easy collect data individuals use some popular models detect modes. However, existing methods are limited aggregation these suggest that travelers' preference mainly depend characteristics...