Run the following command to setup the virtual environment and install the required packages.
make setup
Run the following command to download the Kaggle Landscape dataset with edge images.
make download
Run the following command to generate edges
./env/bin/python3 -m image_processing \
--task hed \
--input_dir DIRECTORY_TO_IMAGES \
--output_dir OUTPUT_DIRECTORY
Then run the following commands in matlab to perform postprocessing
addpath(genpath('toolbox/')); savepath; toolboxCompile;
PostprocessHED(
PATH_TO_HED_EDGES,
OUTPUT_DIRECTORY,
1024, 25.0/255.0, 5
)
Run the following command to perform a test training and evaluation run.
make test
Run the following to download pretrained model
make deploy
Run the following to run the server
./env/bin/python3 app.py
and open http://127.0.0.1:5000/
in the browser.
Simple Edges
- Edges2shoes: https://www.kaggle.com/balraj98/edges2shoes-dataset
- Edges2handbags
Landscape
- (4K) Landscape: https://www.kaggle.com/arnaud58/landscape-pictures
- (90K) Aligning Latent and Image Spaces to Connect the Unconnectable: https://github.com/universome/alis
- Flickr Landscape (need web scrap): https://www.flickr.com/groups/landcape/
Other Style
- Abstract painting: https://www.kaggle.com/flash10042/abstract-paintings-dataset
- Holistically-Nested Edge Detection: https://github.com/s9xie/hed
- Holistically-Nested Edge Detection with OpenCV and Deep Learning: https://www.pyimagesearch.com/2019/03/04/holistically-nested-edge-detection-with-opencv-and-deep-learning/