A5103094528- Robot Manipulation and Learning
- Electrocatalysts for Energy Conversion
- Reinforcement Learning in Robotics
- Advanced battery technologies research
- Advanced Photocatalysis Techniques
- Multimodal Machine Learning Applications
- Robotic Locomotion and Control
- Advanced Vision and Imaging
- Robotic Path Planning Algorithms
- Cell Image Analysis Techniques
- Electrochemical Analysis and Applications
- Multilevel Inverters and Converters
- Fuel Cells and Related Materials
- Catalytic Processes in Materials Science
- Advanced DC-DC Converters
- Anomaly Detection Techniques and Applications
- Human Pose and Action Recognition
- Natural Language Processing Techniques
- Microgrid Control and Optimization
- Advanced Image and Video Retrieval Techniques
- Additive Manufacturing and 3D Printing Technologies
- biodegradable polymer synthesis and properties
- Evolutionary Algorithms and Applications
- Machine Learning and Algorithms
- Bone Tissue Engineering Materials
Beijing Polytechnic
2025
Heilongjiang University
2022-2024
Shandong University
2022-2024
The Ohio State University
2023-2024
Tsinghua University
2024
Google (United States)
2024
DeepMind (United Kingdom)
2024
Sichuan University
2019-2024
Northwestern Polytechnical University
2024
Northeast Forestry University
2022-2024
Integrating metal-based species with a carbon matrix is promising approach for fabricating inexpensive, durable, and efficient electrocatalysts. Herein, NiSe2-decorated N-doped polyhedra (NC) are prepared as electrocatalysts hydrogen evolution reaction (HER) by template-assisted approach. The optimal NC-NiSe2 delivers extraordinary catalytic activities toward HER in wide pH range, overpotentials of 127 mV 0.5 M H2SO4, 226 1 PBS, 205 KOH to drive current density 10 mA cm–2. In addition, this...
Abstract Two-dimensional (2D) superlattices, formed by stacking sublattices of 2D materials, have emerged as a powerful platform for tailoring and enhancing material properties beyond their intrinsic characteristics. However, conventional synthesis methods are limited to pristine sublattices, posing significant practical challenge when it comes chemically modified sublattices. Here we report chemical method that overcomes this creating unique graphene superlattice, with monodisperse,...
Mn-CoP nanosheets were synthesized from ZIF-67 <italic>via</italic> an etching-carbonization–phosphidation strategy and showed efficient electrocatalytic activity in both HER OER under acidic alkaline conditions.
Printed Circuit Board (PCB) is a widely used electronic component and plays critical role in the miniaturization integration of circuits. However, detection PCB defects based on deep learning still encounter difficulties limited efficiency. In order to address issues low speed accuracy defect process, this paper proposed an innovative method YOLOv7. Firstly, FasterNet was applied as backbone network structure. With new partial convolution, spatial features were extracted more efficiently...
Abstract To improve the toughness of polylactic acid (PLA), in this research, a tough polymer polycaprolactone (PCL) was combined with PLA to prepare blends, and compatibilizer glycidyl methacrylate grafted (PLA‐g‐GMA) produced enhance compatibility PLA/PCL blend. Compared blend, compatibilized blend exhibited lower melting temperature greater thermal stability. The tensile test indicates that addition greatly improved Particularly, elongation at break 6 wt% PLA‐g‐GMA increased 370% reached...
The accurate identification of sheep is crucial for breeding, behavioral research, food quality tracking, and disease prevention on modern farms. As a result the time-consuming, expensive, unreliable problems traditional sheep-identification methods, relevant studies have built face recognition models to recognize through facial images. However, existing such as high computational costs, large model sizes, weak practicality. In response above issues, this study proposes lightweight named...
In this work, an automatic machine learning (AutoML) modeling architecture called Autostacker is introduced. combines innovative hierarchical stacking and evolutionary algorithm (EA) to perform efficient parameter search without the need for prior domain knowledge about data or feature preprocessing. Using EA, quickly evolves candidate pipelines with high predictive accuracy. These can be used in their given form, serve as a starting point further augmentation refinement by human experts....
It was aim to study the physicochemical features and biocompatibility of electrospun collagen-chitosan membranes, its potential in guided bone regeneration. Electrospinning technology applied fabrication collagen membranes following observation scanning electron microscope. Physicochemical properties including tensile strength, elongation rate, porosity, degradation were measured then. In vivo, calvarial defects created on rats covered with two kinds respectively. 4th 8th week, enzyme-linked...
This paper focuses on visual semantic navigation, the task of producing actions for an active agent to navigate a specified target object category in unknown environment. To complete this task, algorithm should simultaneously locate and instance category. In comparison traditional point goal requires have stronger contextual prior indoor environments. We introduce SSCNav, that explicitly models scene priors using confidence-aware completion module guide agent's navigation planning. Given...
The surface plasmon resonance (SPR) effect of metal nanoparticles (MNPs) is effectively applied on polymer solar cells (PSCs) to improve power conversion efficiency (PCE). However, universality the reported results mainly focused utilizing single type MNPs enhance light absorption only in specific narrow wavelength range. Herein, a surface-energy-induced dual MNP by thermally evaporating method was presented achieve enhancement wider differences energy between silver (Ag), gold (Au), and...
Abstract In this study, polylactic acid (PLA)/polycaprolactone (PCL)/microcrystalline cellulose (MCC) composites and PLA/PCL/MCC composite compatibilized by the maleic anhydride (MA) copolymer were prepared via two‐screw extrusion technology. The mechanical results showed that strength break elongation of ternary are decreased with increase in MCC contents. After grafting was added, due to better compatibility between components composite, significantly improved compared ungrafted composite....
Abstract Behavior modeling is an essential cognitive ability that underlies many aspects of human and animal social behavior (Watson in Psychol Rev 20:158, 1913), we would like to endow robots. Most studies machine modelling, however, rely on symbolic or selected parametric sensory inputs built-in knowledge relevant a given task. Here, propose observer can model the actor through visual processing alone, without any prior information assumptions about inputs. To test this hypothesis,...