Algorithm which does texture synthesis using convolutional neural networks using two source images and a mask for combination of those sources.
Built on top of this repository - https://github.com/meet-minimalist/Texture-Synthesis-Using-Convolutional-Neural-Networks
This research is in the process of publication in conjuction with other research in Nuclear Forensics at the Scientific Computing and Imaging Institute (SCI) at the University of Utah. Publication will be linked here soon!
Source images to be used are placed in Images folder:
input1_name and input2_name are edited on lines 20 and 21 of main.py to link to the images placed in the Images folder
output_name is edited to specify the desired save name of the output texture image on line 22 of main.py
Outut image is saved into the main project folder.