Xinpeng Li
A5101935387- Advanced Image Processing Techniques
- Image Enhancement Techniques
- Advanced Vision and Imaging
- Quantum Computing Algorithms and Architecture
- Software System Performance and Reliability
- Advanced Image Fusion Techniques
- Domain Adaptation and Few-Shot Learning
- Distributed and Parallel Computing Systems
- Quantum Information and Cryptography
- Advanced Computational Techniques and Applications
- Image Processing Techniques and Applications
- Mobile Agent-Based Network Management
- Robotic Path Planning Algorithms
- Anomaly Detection Techniques and Applications
- Low-power high-performance VLSI design
- Image and Signal Denoising Methods
- Service-Oriented Architecture and Web Services
- Cloud Computing and Resource Management
- Caching and Content Delivery
- Multimodal Machine Learning Applications
- Advancements in Semiconductor Devices and Circuit Design
- Peer-to-Peer Network Technologies
- Advanced Neural Network Applications
- Technology and Data Analysis
- Scheduling and Optimization Algorithms
Case Western Reserve University
2023-2024
First Affiliated Hospital of Xinjiang Medical University
2024
Xinjiang Medical University
2024
Vi Technology (United States)
2021-2023
Megvii (China)
2021-2023
Beijing Institute of Technology
2023
Wuhan Institute of Technology
2023
Beijing Municipal Government
2022
University of Southern Denmark
2022
Huawei Technologies (Sweden)
2022
In this paper, we present an attention-guided deformable convolutional network for hand-held multi-frame high dynamic range (HDR) imaging, namely ADNet. This problem comprises two intractable challenges of how to handle saturation and noise properly tackle misalignments caused by object motion or camera jittering. To address the former, adopt a spatial attention module adaptively select most appropriate regions various expo-sure low (LDR) images fusion. For latter one, propose align...
This paper reviews the NTIRE 2023 challenge on efficient single-image super-resolution with a focus proposed solutions and results. The aim of this is to devise network that reduces one or several aspects such as runtime, parameters, FLOPs, activations, memory footprint, depth RFDN while at least maintaining PSNR 29.00dB DIV2K validation datasets. had 272 registered participants, 35 teams made valid submissions. They gauge state-of-the-art for super-resolution.
This paper reviews the NTIRE 2023 challenge on image denoising (σ = 50) with a focus proposed solutions and results. The aim is to obtain network design capable produce high-quality results best performance measured by PSNR for denoising. Independent additive white Gaussian noise (AWGN) assumed level 50. had 225 registered participants, 16 teams made valid submissions. They gauge state-of-the-art
This work reviews the results of NTIRE 2023 Challenge on Image Shadow Removal. The described set solutions were proposed for a novel dataset, which captures wide range object-light interactions. It consists 1200 roughly pixel aligned pairs real shadow free and affected images, captured in controlled environment. data was white-box setup, using professional equipment lights acquisition sensors. challenge had number 144 participants registered, out 19 teams compared final ranking. extend...
This paper reviews the first challenge on high-dynamic range (HDR) imaging that was part of New Trends in Image Restoration and Enhancement (NTIRE) workshop, held conjunction with CVPR 2021. manuscript focuses newly introduced dataset, proposed methods their results. The aims at estimating a HDR image from one or multiple respective low-dynamic (LDR) observations, which might suffer under-or over-exposed regions different sources noise. is composed by two tracks: In Track 1 only single LDR...
This paper reviews the challenge on constrained high dynamic range (HDR) imaging that was part of New Trends in Image Restoration and Enhancement (NTIRE) workshop, held conjunction with CVPR 2022. manuscript focuses competition set-up, datasets, proposed methods their results. The aims at estimating an HDR image from multiple respective low (LDR) observations, which might suffer under-or over-exposed regions different sources noise. is composed two tracks emphasis fidelity complexity...
Efficient deep learning-based approaches have achieved remarkable performance in single image super-resolution. However, recent studies on efficient super-resolution mainly focused reducing the number of parameters and floating-point operations through various network designs. Although these methods can decrease operations, they may not necessarily reduce actual running time. To address this issue, we propose a novel multi-stage lightweight boosting method, which enable networks to achieve...
High-precision point cloud anomaly detection is the gold standard for identifying defects of advancing machining and precision manufacturing. Despite some methodological advances in this area, scarcity datasets lack a systematic benchmark hinder its development. We introduce Real3D-AD, challenging high-precision dataset, addressing limitations field. With 1,254 high-resolution 3D items from forty thousand to millions points each item, Real3D-AD largest dataset industrial date. surpasses...
Developing and integrating advanced image sensors with novel algorithms in camera systems are prevalent the increasing demand for computational photography imaging on mobile platforms. However, lack of high-quality data research rare opportunity in-depth exchange views from industry academia constrain development intelligent (MIPI). With success 1st MIPI Workshop@ECCV 2022, we introduce second challenge including four tracks focusing algorithms. In this paper, summarize review Nighttime...
With the rapid development of remote-sensing technology and increasing number Earth observation satellites, volume image datasets is growing exponentially. The management big data also becoming increasingly complex difficult, with result that it can be hard for users to access imagery they are interested in quickly, efficiently intelligently. To address these challenges, this paper proposes a image-retrieval model based on an ensemble neural networks. This make full use existing training...
Quantum circuit cutting has been proposed to help execute large quantum circuits using only small and noisy machines. Intuitively, a qubit wire can be thought of as classically passing information state along each element in basis set. As the number cuts increase, degrees freedom needed passed through scales exponentially. We propose simple reduction scheme lower classical resources required perform cut. Particularly, we recognize that for some cuts, certain might pass "no information"...
Rapidly completing the exploration and construction of unknown environments is an important task a UAV cluster. However, formulation online autonomous strategy based on real-time detection map still problem that needs to be discussed optimized. In this paper, we propose distributed environment framework for cluster comprehensively considers path terminal state gain, which called Distributed Next-Best-Path Terminal (DNBPT) method. This method calculates gain by calculating new grid brought...
Disks are the main equipment for data storage in centers. The prediction of disk failure is great significance reliability and security data. On account few abnormal samples datasets, it difficult to satisfy requirement supervised semi-supervised algorithms number while unsupervised have poor performance on recall rate when solving problems local anomalies wrapped a nomalies. This paper presents an incremental learning model using density metric edge samples. An isolation region built by...
Phase change memory (PCM) and DRAM-based hybrid main have emerged as one of the most promising technology for future embedded systems. Previous research works focused on page migration strategies to fully exploit advantages DRAM PCM achieve higher performance energy efficiency. However, ability predict pages’ behavior in these could be further improved do not consider how reduce consumption decrease system energy. In this article, we propose a novel dynamic hardware strategy named periodical...
With the rapid development of computer vision and artificial intelligence, human-computer interaction has become an inevitable part people's lives. Gestures can bring more natural, comfortable, effective communication between people machines. However, in some complex scenarios, such as rooms with looming lighting, robustness universality hand gesture recognition based on traditional cameras are insufficient, supporting algorithms tend to underperform real-time, especially for embedded...
For the planar robot with underactuated constraints, a stable control method is presented on foundation of motion planning and intelligent optimization, which includes two stages. (1) Designing controllers to actuated manipulators given target states. (2) Planning trajectory combined constraints between all links, using algorithm find adaptable parameters, tracking such planned trajectories full states simultaneously. At last, multigroup simulations demonstrate validity proposed method.
Developing and integrating advanced image sensors with novel algorithms in camera systems are prevalent the increasing demand for computational photography imaging on mobile platforms. However, lack of high-quality data research rare opportunity an in-depth exchange views from industry academia constrain development intelligent (MIPI). With success 1st MIPI Workshop@ECCV 2022, we introduce second challenge, including four tracks focusing algorithms. This paper summarizes reviews RGBW Joint...
In order to build large-scale RFID applications, it is necessary establish public service infrastructure, the essential task efficient and secure code resolution network. Aiming at alleviating problems in load balancing, single node failure tolerance security of network, based on related standards research works, a hierarchical P2P network structure proposed, k-Nearest Neighbor storing querying mechanism are presented static resolution, secondary indexing dynamic double random number entity...
In order to build open-loop large scale RFID applications, it is necessary establish public service infrastructure, the key issue efficient and robust code resolution network. But there exist some problems, such as load balancing, single node failure, etc. On basis of recent research, according features governmental administration enterprise a hierarchical P2P based RFDD network structure proposed implemented. The core components are specified formally by TIOA. Meanwhile, management toolset...
RFID technology has become a hot spot in automatic identification filed recent years, which opens up wide variety of applications. However, most applications are still closed-loop-based. In order to build large-scale transregional and cross-industry, it is necessary public service platform. On the basis research, according characteristics enterprise applications, industry administrative aspects, we propose SaaS mode based region platform architecture. The core services encoding. this paper,...