A5101433617- Robotics and Sensor-Based Localization
- Advanced Image and Video Retrieval Techniques
- Anomaly Detection Techniques and Applications
- Software System Performance and Reliability
- Robotic Path Planning Algorithms
- Mobile and Web Applications
- Water Quality Monitoring Technologies
- Indoor and Outdoor Localization Technologies
- Wireless Sensor Networks and IoT
- Software Engineering Research
- Embedded Systems and FPGA Design
- Network Security and Intrusion Detection
- Robotics and Automated Systems
- IoT and GPS-based Vehicle Safety Systems
- Reinforcement Learning in Robotics
- Advanced Data Storage Technologies
- Modular Robots and Swarm Intelligence
- Advanced Neural Network Applications
- 3D Surveying and Cultural Heritage
- Underwater Vehicles and Communication Systems
- Advanced Memory and Neural Computing
- Advanced Decision-Making Techniques
- Robot Manipulation and Learning
- Time Series Analysis and Forecasting
- IoT-based Smart Home Systems
Jingdong (China)
2022-2023
Beijing Jiaotong University
2020-2021
Tsinghua University
2019-2020
North China Electric Power University
2006-2018
Service robots should be able to operate autonomously in dynamic and daily changing environments over an extended period of time. While Simultaneous Localization And Mapping (SLAM) is one the most fundamental problems for robotic autonomy, existing SLAM works are evaluated with data sequences that recorded a short In real-world deployment, there can out-of-sight scene changes caused by both natural factors human activities. For example, home scenarios, objects may movable, replaceable or...
A robust and efficient Simultaneous Localization Mapping (SLAM) system is essential for robot autonomy. For visual SLAM algorithms, though the theoretical framework has been well established most aspects, feature extraction association still empirically designed in cases, can be vulnerable complex environments. This paper shows that with deep convolutional neural networks (CNNs) seamlessly incorporated into a modern framework. The proposed utilizes state-of-the-art CNN to detect keypoints...
Embodied manipulation is a fundamental ability in the realm of embodied artificial intelligence. Although current models show certain generalizations specific settings, they struggle new environments and tasks due to complexity diversity real-world scenarios. The traditional end-to-end data collection training manner leads significant demands, which we call ``data explosion''. To address issue, introduce three-wheeled data-driven method build an atomic skill library. We divide into subtasks...
Unmanned Aerial Vehicles (UAVs) can significantly improve the autonomy of mining industry, and self-localization is key to autonomous flights underground UAVs. A localization method visual-inertial sensor data fusion proposed in this paper. The aims accuracy robustness UAVs dynamic environments. First, an algorithm for point detection rejection presented, which combines a semantic segmentation neural network, optical flow method, epipolar constraint method. Second, used enhance performance...
In order to reduce the time and space complexity, this paper presents a study on key generation of RSA public-key cryptosystem. An improved prescreening algorithm is introduced make screen for large random number which generated by generator. After that, Miller-Rabin used final prime test. At last, Stain generate public private keys. The results show that research effectively efficiency generation.
As Embodied AI advances, it increasingly enables robots to handle the complexity of household manipulation tasks more effectively. However, application in these settings remains limited due scarcity bimanual-mobile robot datasets. Existing datasets either focus solely on simple grasping using single-arm without mobility, or collect sensor data a narrow scope sensory inputs. result, often fail encapsulate intricate and dynamic nature real-world that are expected perform. To address...
Log data is a valuable resource for understanding system status. recording running status computer commonly used to identify performance issues and malfunctions. Sequential anomaly detection of logs crucial building secure stable beneficial the discovery, location, analysis failures. In this paper, we propose new log sequential method based on natural language processing techniques by Population Based Training (PBT) algorithm, which can make full use semantic information in templates analyze...
Service robots should be able to operate autonomously in dynamic and daily changing environments over an extended period of time. While Simultaneous Localization And Mapping (SLAM) is one the most fundamental problems for robotic autonomy, existing SLAM works are evaluated with data sequences that recorded a short In real-world deployment, there can out-of-sight scene changes caused by both natural factors human activities. For example, home scenarios, objects may movable, replaceable or...
Feature extraction plays an important role in visual localization. Unreliable features on dynamic objects or repetitive regions will interfere with feature matching and challenge indoor localization greatly. To address the problem, we propose a novel network, RaP-Net, to simultaneously predict region-wise invariability point-wise reliability, then extract by considering both of them. We also introduce new dataset, named OpenLORIS-Location, train proposed network. The dataset contains 1553...
System logs are widely used by engineers to record runtime status in the information technology (IT) field. The sequential anomaly detection of is crucial for building a secure and stable system beneficial discovery, location, analysis failures. Conventional manual log suffers high costs unsustainable development. Thus, automatic methods based on Natural Language Processing (NLP) proposed improve accuracy efficiency detection. In this paper, we propose new model, named LogPS. LogPS utilizes...
This paper presents a new design of vehicle remote monitoring system. ARM9 core processor, GPRS module, GPS sensor module and display are used to build car terminal. And develops piece software that runs on center server with .NET development. Based these, client based android platform is developed help user manage their car. The whole system consists four parts: terminal, server, smart phone application wireless communication network. can real-time views information by using at any time...
Offline reinforcement learning (RL) aims to learn optimal policies from previously collected datasets. Recently, due their powerful representational capabilities, diffusion models have shown significant potential as policy for offline RL issues. However, previous algorithms based on generally adopt weighted regression improve the policy. This approach optimizes only using actions and is sensitive Q-values, which limits further performance enhancement. To this end, we propose a novel...
This article describes a vehicle monitoring system client based on HTML and ASP.NET. It involves the technology, ASP.NET etc. with excellent compatibility expansibility. Their key functions are sort cars according to their colors, query list, state information display, data report, terminal controls, According distribution running of vehicles achieve reasonable dispatch, monitor vehicles. will increase efficiency, benefit safety whole system. The has broad development prospect great market...
To achieve real-time monitor and control of the water supply remote monitoring terminal management other information(data equipment); this paper introduces design a system for pipe network with .NET development platform, on basis various key technologies such as GIS, database, data communication, multi-threaded etc. The adopts multi-layer architecture framework to scientific, dynamic visual well. By running software on-site verifies reasonableness stability operation provides feasible...
The log data that records the operating state of a computer system is great significance for understanding state. Log classification crucial engineers to monitor running status and analysis failures. To improve representation quality template reduce model inference time, we propose new method based on natural language processing techniques. In this paper, three embedding methods are adopted complete word vectorization process digital templates, which can make full use semantic information,...
A robust and efficient Simultaneous Localization Mapping (SLAM) system is essential for robot autonomy. For visual SLAM algorithms, though the theoretical framework has been well established most aspects, feature extraction association still empirically designed in cases, can be vulnerable complex environments. This paper shows that with deep convolutional neural networks (CNNs) seamlessly incorporated into a modern framework. The proposed utilizes state-of-the-art CNN to detect keypoints...
With the swift implementation of cloud platforms, guaranteeing optimal service dependability is imperative. Disk failure a common cause unreliability, particularly in industrial platforms that house numerous disks. These approaches can forecast disk failures by analyzing status data before they occur. Deep neural networks have demonstrated their effectiveness addressing time series classification problems. Our research paper introduces an innovative and efficient approach to forecasting...
With the rapid development of Smart city, water is an important part which paid more and attention. It obtains deals with urban information through technology. can effectively manage supply, The sale other processes. At same time, due to popularity Smartphones, Smartphone applications have covered every aspect life become indispensable people's daily life. Through applications, user achieve online mobile purchase, query situation, quality basic greatly facilitate use user, for wisdom...
Feature extraction plays an important role in visual localization. Unreliable features on dynamic objects or repetitive regions will interfere with feature matching and challenge indoor localization greatly. To address the problem, we propose a novel network, RaP-Net, to simultaneously predict region-wise invariability point-wise reliability, then extract by considering both of them. We also introduce new dataset, named OpenLORIS-Location, train proposed network. The dataset contains 1553...