Jiafu Chen

ORCID: 0000-0001-8253-2809
Publications
Citations
Views
---
Saved
---
About
Contact & Profiles
Research Areas
  • Advanced Image Processing Techniques
  • Supercapacitor Materials and Fabrication
  • Computer Graphics and Visualization Techniques
  • Photochemistry and Electron Transfer Studies
  • Orbital Angular Momentum in Optics
  • Generative Adversarial Networks and Image Synthesis
  • Advanced Vision and Imaging
  • Free Radicals and Antioxidants
  • Advanced battery technologies research
  • Concrete and Cement Materials Research
  • Civil and Geotechnical Engineering Research
  • Electrocatalysts for Energy Conversion
  • Optical Network Technologies
  • Natural Language Processing Techniques
  • Image Enhancement Techniques
  • Advanced Neural Network Applications
  • Advanced Image and Video Retrieval Techniques
  • Metamaterials and Metasurfaces Applications
  • Spectroscopy and Quantum Chemical Studies
  • Neural Networks and Reservoir Computing
  • Photonic and Optical Devices
  • Chalcogenide Semiconductor Thin Films
  • Optical Polarization and Ellipsometry
  • Perovskite Materials and Applications
  • Advanced Chemical Physics Studies

Southern Medical University
2025

Nanfang Hospital
2025

Northwestern Polytechnical University
2023-2024

Shenzhen University
2022-2024

Zhejiang University
2023-2024

Shenzhen Technology University
2023

Zhejiang University of Science and Technology
2023

Shanghai University of Electric Power
2023

Jiangsu Provincial Meteorological Bureau
2023

University of Science and Technology of China
2004-2020

Abstract A photothermally sensitive poly( N ‐isopropylacrylamide)/graphene oxide (PNIPAM/GO) nanocomposite hydrogel can be synthesized by in situ γ‐irradiation‐assisted polymerization of an aqueous solution ‐isopropylacrylamide monomer the presence graphene (GO). The colors and phase‐transition temperatures PNIPAM/GO hydrogels change with different GO doping levels. Due to high optical absorbance GO, shows excellent photothermal properties, where its phase transitions controlled remotely...

10.1002/adfm.201201020 article EN Advanced Functional Materials 2012-06-12

Nanostructured carbon aerogels with outstanding physicochemical properties have exhibited great application potentials in widespread fields and therefore attracted extensive attentions recently. It is still a challenge so far to develop flexible economical routes fabricate high‐performance nanocarbon aerogels, preferably based on renewable resources. Here, ultralight multifunctional reduced graphene oxide/carbon nanofiber (RGO/CNF) are fabricated from oxide low‐cost, industrially produced...

10.1002/smll.201700453 article EN Small 2017-05-16

Multifunctional nanocomposite hydrogel: swelling-shrink transition of the magnetic sensitive poly(N-isopropylacrylamide)/Fe3O4 (PNIPAM/Fe3O4) hydrogel can be controlled via near-infrared (NIR) laser exposure or non-exposure, which shows potential as a movable position heating source manipulated by combination an external magnet and irradiation.

10.1002/smll.201400477 article EN Small 2014-03-27

This paper presents a new adversarial training framework for image inpainting with segmentation confusion (SCAT) and contrastive learning. SCAT plays an game between generator network, which provides pixel-level local signals can adapt to images free-form holes. By combining standard global training, the exhibits following three advantages simultaneously: (1) consistency of repaired image, (2) fine texture details (3) flexibility handling Moreover, we propose textural semantic learning...

10.1609/aaai.v37i3.25502 article EN Proceedings of the AAAI Conference on Artificial Intelligence 2023-06-26

Recently proposed LaMa [25] introduce Fast Fourier Convolution (FFC) [4] into image inpainting. FFC empowers the fully convolutional network to have a global receptive field in its early layers, and ability produce robust repeating texture. However, has difficulty generating clear sharp complex content. In this paper, we analyze fundamental flaws of using inpainting, which are 1) spectrum shifting, 2) unexpected spatial activation, 3) limited frequency field. Such make FFC-based inpainting...

10.1109/iccv51070.2023.02120 article EN 2021 IEEE/CVF International Conference on Computer Vision (ICCV) 2023-10-01

Abstract Affording the unique attributes of parallel processing, high-speed operation, and multi-dimensional manipulation, optical encryption has garnered attention as an excellent candidate for high-security information encoding decoding. However, existing schemes that employ common photonic dimensions like wavelength, polarization, angular momentum are subject to upper limit O(2N), significantly hampering development security. To overcome this limitation, we present a novel approach, which...

10.1088/2040-8986/adc820 article EN Journal of Optics 2025-04-02

Recently, several methods have explored the potential of multi-contrast magnetic resonance imaging (MRI) super-resolution (SR) and obtain results superior to single-contrast SR methods. However, existing approaches still two shortcomings: (1) They can only address fixed integer upsampling scales, such as 2×, 3×, 4×, which require training storing corresponding model separately for each scale in clinic. (2) lack direct interaction among different windows they adopt square window (e.g., 8×8)...

10.1109/iccv51070.2023.01941 article EN 2021 IEEE/CVF International Conference on Computer Vision (ICCV) 2023-10-01

Text-driven 3D style transfer aims at stylizing a scene according to the text and generating arbitrary novel views with consistency. Simply combining image/video methods view synthesis results in flickering when changing viewpoints, while existing learn styles from images instead of texts. To address this problem, we for first time design an efficient text-driven model transfer, named TeSTNeRF, stylize using texts via cross-modal learning: leverage advanced encoder embed order control align...

10.24963/ijcai.2023/642 article EN 2023-08-01

Neural Radiance Fields (NeRF) has shown great success in novel view synthesis due to its state-of-the-art quality and flexibility. However, NeRF requires dense input views (tens hundreds) a long training time (hours days) for single scene generate high-fidelity images. Although using the voxel grids represent radiance field can significantly accelerate optimization process, we observe that sparse inputs, are more prone overfitting will have holes floaters, which leads artifacts. In this...

10.24963/ijcai.2023/157 article EN 2023-08-01

To address the problem of insufficient semantic feature information caused by small objects in aerial images, an adaptive receptive field enhancement network based on attention mechanism suitable for scene is proposed. First, to make model more deployment UAV platforms with limited resources, block reconstructed a series branch multiplexing, which can output map front pass through convolutional layer back further extraction, improving utilization resources. Second, expand as little loss...

10.1109/tgrs.2023.3337266 article EN IEEE Transactions on Geoscience and Remote Sensing 2023-11-27

Cylindrical vector beams (CVBs) exhibit great potential for multiplexing communication, owing to their mode orthogonality and compatibility with conventional wavelength techniques. However, the practical application of CVB communication faces challenges due lack effective spatial polarization manipulation technologies (de)multiplexing multi-dimensional physical dimensions CVBs. Herein, we introduce a wavelength- polarization-sensitive cascaded phase modulation strategy that utilizes multiple...

10.1364/oe.514218 article EN cc-by Optics Express 2024-02-21
Coming Soon ...