A5050531685- Simulation and Modeling Applications
- Industrial Vision Systems and Defect Detection
- Image Processing Techniques and Applications
- Image and Object Detection Techniques
- Advanced Neural Network Applications
- Acoustic Wave Phenomena Research
- Advanced Vision and Imaging
- Optical measurement and interference techniques
- Aerodynamics and Acoustics in Jet Flows
- Microfluidic and Bio-sensing Technologies
- Sports Dynamics and Biomechanics
- Speech and Audio Processing
- Traffic Prediction and Management Techniques
- Robotic Mechanisms and Dynamics
- Remote Sensing and LiDAR Applications
- Transportation Planning and Optimization
- Computer Graphics and Visualization Techniques
- Human Pose and Action Recognition
- Underwater Acoustics Research
- Adaptive Control of Nonlinear Systems
- Advanced Image and Video Retrieval Techniques
- Robotic Path Planning Algorithms
- Surface Roughness and Optical Measurements
- Higher Education and Teaching Methods
- Iterative Learning Control Systems
Chinese Academy of Sciences
2011-2025
Shandong Institute of Automation
2012-2025
Institute of Automation
2011-2024
ShanghaiTech University
2024
University of Electronic Science and Technology of China
2023
Southwest Jiaotong University
2007-2023
University of Chinese Academy of Sciences
2017-2023
Beijing Jiaotong University
2021
Vision Technology (United States)
2018-2019
Sun Yat-sen University
2007-2015
Moving object segmentation (MOS), aiming at segmenting moving objects from video frames, is an important and challenging task in computer vision with various applications. With the development of deep learning (DL), MOS has also entered era models toward spatiotemporal feature learning. This paper aims to provide latest review recent DL-based methods proposed during past three years. Specifically, we present a more up-to-date categorization based on model characteristics, then compare...
Acoustic holography is a promising technique for contactless manipulation, remote sensing, and energy harvesting. It involves retrieving holograms used to modulate acoustic sources reconstructing target fields. The performance of reconstruction primarily determined by two key criteria, including the spatial bandwidth product, which measures pixel number representing information capacity, resolution, quantifies size supporting detail gain. However, existing techniques face limitations in...
Underwater object detection presents both significant challenges and opportunities within ocean exploration conservation. Although the current popular algorithms generally achieve strong performance. Because underwater images are affected by insufficient illumination, wavelength-dependent scattering, absorption, performance for objects is suboptimal. Therefore, a local channel information encoding method named Partial Semantic Encoding Module (PSEM) an attention based head called Split...
Recently, many different types of artificial neural networks (ANNs) have been applied to forecast stock price and good performance is obtained. However, most these models use only a small number features as input there may not be enough information make prediction due the complexity market. If having larger features, run time training would increased generalization deteriorated curse dimension. Therefore, an effective tool extract highly discriminative low-dimensional from high-dimensional...
An analytic flying model that can well represent the physical behavior is derived, where ball's self-rotational velocity changes along with velocity. Based on least square method, a rebound represents relation between velocities before and after established. The initial trajectory fitted to three second order polynomials of time measured positions ball. ball in model, including velocity, are computed from polynomials. landing position predicted model. determined By taking as new ones,...
In this paper, a novel prediction method of the striking position is proposed for robotic ping-pong player. order to remove noise involved in coordinates ball, new nonlinear filter based on fuzzy logic approach presented. Then, least square (LSM) utilized compute initial flying and rotational velocities filtered positions ball. The impact between ball table studied, analytic model that represents relationship before after rebound developed. Based statistical analysis, memory-based local...
In this paper, an efficient insertion control strategy is presented for precision assembly of cylindrical components with interference fit. It has two inherent advantages compared existing methods in the domain assembly. First, it does not need preparatory actions dedicated horizontal forces before actions, which realized by integrating preparations into step. Second, instead inserting fixed incremental depth, estimates depth every action to be as large possible a probabilistic approach....
Both accuracy and efficiency are significant for pose estimation tracking in videos. State-of-the-art performance is dominated by two-stages top-down methods. Despite the leading results, these methods impractical real-world applications due to their separated architectures complicated calculation. This paper addresses task of articulated multi-person towards real-time speed. An end-to-end multi-task network (MTN) designed perform human detection, estimation, person re-identification (Re-ID)...
The installation of a ground energy storage system (ESS) in the substation can improve recovery and utilization regenerative braking energy. This paper proposes an management strategy (EMS) adaptive threshold adjustment for ESS. In this regard, analyzes flow traction power supply (TPSS) with different headways no-load voltage. According to influence charge discharge on TPSS, laws are formulated. Subsequently, EMS based fuzzy logic control (FLC) is designed. proposed variation headway By...
Acoustic holography (AH) provides a promising technique for arbitrary acoustic field reconstruction, supporting many applications like robotic micro-nano manipulation, neuromodulation, volumetric imaging, and virtual reality. In AH, three-dimensional (3D) fields quantified with complex-valued pressures are reconstructed by virtue of two-dimensional (2D) holograms. Phase-only hologram (POH) is recently regarded as an energy-efficient way which typically implemented dynamically programmable...
Deep-learning-based detection methods have been widely applied to industrial defect inspection. However, directly using vanilla fails achieve satisfying performance due the lack of identifiable features. In this paper, a novel attention-based multi-scale feature fusion method (AMFF) is proposed, aiming enhance features and improve identification by leveraging attention mechanism in fusion. AMFF includes self-enhanced module (SEAM) cross-enhanced (CEAM). SEAM performed on single map, which...
Audio-Visual Source Localization (AVSL) is the task of identifying specific sounding objects in scene given audio cues. In our work, we focus on semi-supervised AVSL with pseudo-labeling. To address issues vanilla hard pseudo-labels including bias accumulation, noise sensitivity, and instability, propose a novel method named Cross Pseudo-Labeling (XPL), wherein two models learn from each other cross-refine mechanism to avoid accumulation. We equip XPL effective components. Firstly, soft...
In this brief, we consider the problem of controlling racket attached to ping-pong playing robot, so that incoming ball is returned a desired position. The maps are used calculate racket's initial parameters described. They implemented with locally weighted regression (LWR). An active learning approach based on fuzzy cerebellar model articulation controller (FCMAC) proposed, and then it added LWR, which regarded as lazy learning. A algorithm for updating experience data in CMAC according...