F
Feng Yuchao
Researcher at Zhejiang University of Technology
Publications - 5
Citations - 122
Feng Yuchao is an academic researcher from Zhejiang University of Technology. The author has contributed to research in topics: Convolution & Feature (computer vision). The author has an hindex of 2, co-authored 5 publications receiving 18 citations.
Papers
More filters
Journal ArticleDOI
Hyperspectral Image Classification Using Mixed Convolutions and Covariance Pooling
TL;DR: This work proposes a mixed CNN with covariance pooling for HSI classification that starts with spectral-spatial 3-D convolutions that followed by a spatial 2-D Convolution, which achieves better accuracy than other state-of-the-art methods.
Journal ArticleDOI
Fast Tensor Nuclear Norm for Structured Low-Rank Visual Inpainting
TL;DR: This paper develops a new Hankel low-rank tensor recovery method that is competent to truthfully capture the underlying details with sacrifice of slightly more computational burden and proposes a fast randomized skinny tensor singular value decomposition (rst-SVD) to accelerate the per-iteration running efficiency.
Journal ArticleDOI
3D Octave and 2D Vanilla Mixed Convolutional Neural Network for Hyperspectral Image Classification with Limited Samples
TL;DR: Wang et al. as mentioned in this paper designed an end-to-end 3D octave and 2D vanilla mixed CNN, namely Oct-MCNN-HS, based on the typical 3D-2D mixed CNN (MCNN).
Journal ArticleDOI
Tensor completion using patch-wise high order Hankelization and randomized tensor ring initialization
TL;DR: Wang et al. as mentioned in this paper provided a detailed analysis on the rank properties in patch-wise and Hankel cases, and designed a resultful yet simple tensor completion method using patchwise and multi-way tensor extension.
Patent
Hyperspectral remote sensing image classification method based on three-dimensional and two-dimensional hybrid convolution
TL;DR: In this article, a hyperspectral remote sensing image classification method based on three-dimensional and two-dimensional hybrid convolution was proposed, which combines the advantages of 3D convolution and 2D convolutions and achieves the accurate classification of HRS images under low training samples.