- Robot Manipulation and Learning
- Robotic Path Planning Algorithms
- Photonic and Optical Devices
- Robotics and Sensor-Based Localization
- Advanced Image and Video Retrieval Techniques
- Vehicle License Plate Recognition
- Data Management and Algorithms
- Robotic Locomotion and Control
- Handwritten Text Recognition Techniques
- Image Retrieval and Classification Techniques
- Soft Robotics and Applications
- Advanced Welding Techniques Analysis
- Muscle activation and electromyography studies
- Image Processing and 3D Reconstruction
- Advanced Fiber Laser Technologies
- Terahertz technology and applications
- Acoustic Wave Resonator Technologies
- Hand Gesture Recognition Systems
- 3D Surveying and Cultural Heritage
- Aluminum Alloys Composites Properties
- Ferroelectric and Piezoelectric Materials
- Aerospace and Aviation Technology
- Advanced Photocatalysis Techniques
- Currency Recognition and Detection
- Reinforcement Learning in Robotics
Sinopec (China)
2024
Qingdao University of Science and Technology
2024
Harbin University of Commerce
2022-2023
Huazhong University of Science and Technology
2007-2023
Fudan University
1999-2021
University of Jinan
2018
China University of Mining and Technology
2015-2017
Hsin Sheng College of Medical Care and Management
2008
Wayne State University
1996-2002
Materials Modification (United States)
1999
Through two theoretical lenses, the Health Belief Model (HBM) and literature explaining conspiracy theories, this study examined media discourse about Human Papillomavirus (HPV) vaccine in China's cyberspace. Results showed that narratives are positive but misleading, HBM components presented inconsistently over time. The also identified emerging theories HPV vaccine: Chinese theorists accused Western countries of using to destroy ethnic group. These findings demonstrate influence public...
The phosphate adsorption mechanisms of SA-0.4La@De involved electrostatic interactions, ligand exchange, and the formation La–P complexes.
We perform a search of double beta decay $^{136}$Xe to the excited state, $0^+_1$, $^{136}$Ba (2$\nu \beta \beta$-0$_1^+$), using dual-phase xenon detector PandaX-4T with first 94.9-day commissioning data. The multi-site events are reconstructed up MeV energy scale, which helps improve background model significantly. contribution from stainless steel platform outside cryostat is evaluated for time. No significant evidence 2$\nu \beta$-0$_1^+$ observed, resulting in lower limit $T_{1/2}^{2...
As a new type of promising semiconductor photocatalyst, SnO2 cannot be widely applied due to its low utilization efficiency visible light and swift recombination photogenerated electrons holes. These drawbacks were effectively overcome by preparing the B/Ag/F tridoped SnO2-ZnO composite films using simple sol–gel method. The degradation methyl green formaldehyde solutions was used value photocatalytic activity samples. Photoluminescence (PL) spectra UV–Vis absorption spectroscopy results...
Aqueous gelcasting for 0.94Bi 0.5 Na TiO 3 –0.06BaTiO (BNBT6) lead‐free piezoceramics was investigated in this paper. A stable BNBT6 suspension with 50 vol% of solid loading and <1Pa·s viscosity successfully prepared when wt% PMAA‐NH 4 added the pH value controlled range 8–10. Dehydration desiccation gelled parts could be easily through dipping pretreatment a 5 ammonium persulfate solution 24 h or 10 12 h. Scanning electron microscopy micrographs show that ceramic by route exhibits...
The segmentation of off-line cursive handwriting is an important step in recognition. In this paper, a new approach described for task. suggested uses set heuristic rules to determine possible letter boundaries the image word. are based on associations that exist between certain geometric and topologic features English language characters. A system incorporating proposed has been built perform postal address images. includes several preprocessing steps extract handwritten words from envelope...
This paper studies the task of image captioning with novel objects, which only exist in testing images. Intrinsically, this can reflect generalization ability models understanding and semantic meanings visual concepts objects unseen training set, sharing similarity to one/zero-shot learning. The critical difficulty thus comes from that no paired images sentences be used help train model. Inspired by recent work (Chen et al. 2019b) boosts one-shot learning generate various deformations, we...
Abstract Predicting the activity of solar flares is great significance for studying its physical mechanism and impact on human production life. Problems such as class imbalance, high time-series sensitivity, over-localization important features exist in sample data used flare forecasting. We design a fusion method based resampling CNN-GRU algorithm to try solve above problems. In order verify effectiveness this method, first, we compared forecast performance different methods by keeping...
Abstract This study presents a generalized version of the residual differential item functioning (RDIF) detection framework in response theory, named GRDIF, to analyze (DIF) multiple groups. The GRDIF retains advantages original RDIF framework, such as computational efficiency and ease implementation. performance was assessed through simulation compared with existing DIF methods, including Mantel‐Haenszel, Lasso‐DIF, alignment methods. Results showed that demonstrated well‐controlled Type I...
Establishing the identify of a signature by automatic search through database signatures is interest in several areas. This paper describes system for this purpose. The proposed identification uses set geometric and topologic features to characterize each signature. By considering spatial distribution these features, maps into two strings finite symbols. A local associative indexing scheme then used on organize collection known identity. When presented with unknown identity, same retrieve...
Off-line cursive script recognition has got increasing attention during the last three decades since it is of interest in several areas such as banking and postal service. An off-line handwritten word system described this paper used for legal amount interpretation personal checks. The proposed uses a set geometric topologic features to characterize each word. By considering spatial distribution these image, maps into two strings finite symbols. A local associative indexing scheme then on...
Abstract A method is introduced to predict human motion trajectory in the process of human-robot collaboration (HRC). In method, distances are assumed be a Gaussian Process (GP). To achieve this, handover task conducted by and collaborative robot, while positions hand robot end-effector recorded. Some recorded data used for Regression (GPR), GP 95% confidence convince about obtained GPR. Experimental results show that 80% testing included convince. The here useful other tasks where existing...
We propose and demonstrate an individually controllable TE–TM polarization splitter–combiner. Using these characteristics, we can tune the cross-talk values of fabricated devices to their optimum values. The measured for TE TM polarizations our device are lower than −20 dB at operational voltages less 20 V.
The influences of the substitution (Zn1/3Nb2/3) group for Ti as well addition Fe2O3 and K2CO3 on d33, kp, εr, tgδ Pb[Zr0.46Ti0.54-x(Zn1/3Nb2/3)x]O3 + y mol%Fe2O3+z mol%K2CO3 (PZNPZT) (0 ≤ x 0.20, 0 0.10, z 0.20) ceramics were investigated. experimental results demonstrated that when increases in Pb[Zr0.46Ti0.54-x(Zn1/3Nb2/3)x]O3+0.3 mol%Fe2O3+0.5mol%K2CO3 ceramics, increase approach to their maximum values, then eventually decrease. XRD analysis revealed perovskite structure materials...
With the rapid development of Industrial Internet Things (IIoT), prediction Remaining Useful Life (RUL) has played a crucial role in ensuring equipment reliability, enhancing production efficiency, and reducing maintenance costs. Particularly today's data-driven industrial environment, adoption effective RUL techniques is paramount. In response to this demand, we propose implement deep learning-based model. Our model begins by utilizing one-dimensional Convolutional Neural Network (CNN)...
We report the results of organic geochemical analyses 19 crude oil samples from reservoir sandstones in 4th Member Eocene Shahejie Formation wells Minfeng Sag, Dongying Depression, Bohai Bay Basin (NE China). In addition, 42 core dark‐coloured mudstones, including 28 extracts, were analysed. Geochemical data included Rock‐Eval measurements, gas chromatography, GC‐MS and diamondoid analyses. Maceral showed that mudstones (“Es 4 ”) contain Types I II 1 kerogen. The member can be divided into...
When the transformer is energized, three-phase inrush current with zero-mode component named (ZMIC) will be generated due to inconsistent saturation of iron core. From on-site waveform record, ZMIC attenuates very slowly and leads zero-sequence zone Ⅲ backup protection malfunction upstream transmission line transformer. The expands outages has an adverse effect on equipment safety system stability. To solve this problem, analytical model under sympathetic interaction considering residual...
Convolutional neural networks for pose estimation based on 2D images are severely affected by imaging quality, illumination changes and occlusion, which greatly reduces their reliability robustness in cluttered scenes. Hence a point cloud convolutional network orientation estimation(OCNN) is proposed this paper. With the network, 3D estimated regressing to rotation matrix. An orthogonality metric loss introduced normalize output For objects with rotational symmetry, an algorithm estimating...
Polarization splitters that use an asymmetric branching waveguide and can be fabricated along the Z axis as well X of LiNbO(3) are proposed demonstrated. In these devices, two polarization states converted to different spatial modes, modes split by waveguide. The extinction ratios shown less than -12 dB.