Ji Liu

ORCID: 0000-0002-4703-991X
Publications
Citations
Views
---
Saved
---
About
Contact & Profiles
Research Areas
  • Astrophysics and Cosmic Phenomena
  • Advanced Antenna and Metasurface Technologies
  • Antenna Design and Optimization
  • Dark Matter and Cosmic Phenomena
  • Electromagnetic Scattering and Analysis
  • Neural Networks and Reservoir Computing
  • Domain Adaptation and Few-Shot Learning
  • Radio Astronomy Observations and Technology
  • Model Reduction and Neural Networks
  • Force Microscopy Techniques and Applications
  • Advanced SAR Imaging Techniques
  • Radio Wave Propagation Studies
  • Neutrino Physics Research
  • Ionosphere and magnetosphere dynamics
  • Neural Networks and Applications
  • Gene Regulatory Network Analysis
  • Ship Hydrodynamics and Maneuverability
  • Mathematical Biology Tumor Growth
  • Traditional Chinese Medicine Studies
  • Ophthalmology and Visual Impairment Studies
  • Corneal surgery and disorders
  • Solar and Space Plasma Dynamics
  • Advanced Measurement and Detection Methods
  • Glaucoma and retinal disorders
  • Advanced Computational Techniques and Applications

University of Alberta
2024

Kuaishou (China)
2021

Beihang University
2015-2020

Yale University
2015

Air Force General Hospital PLA
2015

People's Hospital of Shiyan
2014

Chinese Academy of Sciences
2014

Northwest Normal University
2013

We propose ResRep, a novel method for lossless channel pruning (a.k.a. filter pruning), which slims down CNN by reducing the width (number of output channels) convolutional layers. Inspired neurobiology research about independence remembering and forgetting, we to re-parameterize into parts forgetting parts, where former learn maintain performance latter prune. Via training with regular SGD on but update rule penalty gradients latter, realize structured sparsity. Then equivalently merge...

10.1109/iccv48922.2021.00447 article EN 2021 IEEE/CVF International Conference on Computer Vision (ICCV) 2021-10-01

Designing an efficient and effective neural network has remained a prominent topic in computer vision research. Depthwise onvolution (DWConv) is widely used CNNs or ViTs, but it needs frequent memory access during inference, which leads to low throughput. FasterNet attempts introduce partial convolution (PConv) as alternative DWConv compromises the accuracy due underutilized channels. To remedy this shortcoming consider redundancy between feature map channels, we novel Partial visual...

10.32388/1l3te6 preprint EN cc-by 2025-03-21

Model compression techniques are recently gaining explosive attention for obtaining efficient AI models various real time applications. Channel pruning is one important strategy, and widely used in slimming DNNs. Previous gate-based or importance-based methods aim to remove channels whose "importance" smallest. However, it remains unclear what criteria the channel importance should be measured on, leading selection heuristics. Some other sampling-based deploy sampling strategy train...

10.1109/iccv48922.2021.00519 article EN 2021 IEEE/CVF International Conference on Computer Vision (ICCV) 2021-10-01

This paper deals with a Neumann boundary value problem for Keller-Segel model cubic source term in d -dimensional box<mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML" id="M1"><mml:mrow><mml:mrow><mml:mo stretchy="false">(</mml:mo><mml:mrow><mml:mi>d</mml:mi><mml:mo>=</mml:mo><mml:mn>1</mml:mn><mml:mo>,</mml:mo><mml:mn>2</mml:mn><mml:mo>,</mml:mo><mml:mn>3</mml:mn></mml:mrow><mml:mo stretchy="false">)</mml:mo></mml:mrow></mml:mrow></mml:math>, which describes the movement of cells...

10.1155/2013/934745 article EN cc-by Advances in Mathematical Physics 2013-01-01

<title>Abstract</title> In regions of the solar system distant from planetary magnetic fields, galactic cosmic rays (GCRs) have generally been assumed to be uniformly distributed throughout interplanetary space. However, our analysis data Chang’E-4 lander revealed a region reduced GCR flux in pre-noon sector lunar orbit. Further investigation suggests presence an energetic particle cavity, formed by Earth’s field acting as obstacle propagation. This cavity indicates that influence within...

10.21203/rs.3.rs-5447254/v1 preprint EN cc-by Research Square (Research Square) 2024-12-12

This paper studies radar cross section of submarine sails on the water. Under detection enemy’s airborne radar, considering motion submarine, sail model pitch angle incidence range analysis is established. By using CATIA software, 3D models AKULA sail, SUBOFF and VICTOR are built. On basis physical optics method equivalent currents method, scattering characteristics RCS(radar section) simulated under X band. Through microwave anechoic chamber test, this verifies accuracy combination...

10.1051/itmconf/20181702003 article EN cc-by ITM Web of Conferences 2018-01-01

Abstract Airfoil is one of the strong scattering sources aircraft, so it great significance to study stealth characteristics airfoil for reduction radar cross section whole aircraft. As known, aircraft main component that generates lift. Shaping will reduce aerodynamic performance However, emergence metamaterials allows designers achieve a good balance between and aerodynamics. Considering high cost test, numerical simulation proposed guide design metamaterial stealth. Then, bistatic (RCS)...

10.1088/1742-6596/1633/1/012019 article EN Journal of Physics Conference Series 2020-09-01

The NuMI off-Axis $\nu_e$ Appearance (NO$\nu$A) experiment is a long baseline, off-axis neutrino oscillation experiment. It designed to search for oscillations of $\nu_\mu$ by comparing measurements the beam composition in two detectors. These detectors are functionally identical, nearly fully-active liquid-scintillator tracking calorimeters and located at points along line observe neutrinos. Near Detector (ND), situated \unit[1]{km} away from proton target Fermilab, measures neutrinos prior...

10.21220/s2gh2b article EN 2016-01-01
Coming Soon ...