zhaomin0101 / 3dcnn-hu Goto Github PK

Implementation of paper "Hyperspectral Unmixing for Additive Nonlinear Models With a 3-D-CNN Autoencoder Network" (T-GRS 2022)

Home Page: https://ieeexplore.ieee.org/document/9503107

License: MIT License

MATLAB 5.71% Python 94.29%

3dcnn-hu's Issues

数据形状

Hello, may I ask why the input of the endmember is 468×3 and the original endmember is not 156×3?

端元为负

请问对所给参考进行可视化，发现端元为负数？

What parameters need to be changed and how.

Hello
I am a novice in this field. I've been trying to debug your program recently. If I have my own dataset, can you please tell me what parameters and code I need to change?
Thank you.

Datashape problem

Hi Zhamoin.
Your work is so interesting, you did a great job. I am trying to implement your work, but i am facing small problem. why you changed intput data into

hsi.shape
Out[638]: (9025, 9, 156)

how 9 columns are needed. Please help me with this.

I really appreciate any help you can provide.

A.Suresh.

Channel Kernel size of 3D conv

Hello, i am wondering how to define the channel filter kernel of the five 3D conv layers in the encoder for different datasets, such as Urban, Samson, Houston, JasperRidge?

Recommend Projects