pytorch-hed 0.5

pytorch-hed - Holistically-Nested Edge Detection based on Pytorch

Pytorch Holistically-Nested Edge Detection (HED)

• Website: davidelanz.github.io/pytorch-hed
• Documentation: pytorch-hed.readthedocs.io

This is a reimplementation in the form of a python package of Holistically-Nested Edge Detection [1] using PyTorch based on the previous pytorch implementation by sniklaus [2]. If you would like to use of this work, please cite the paper accordingly. Also, make sure to adhere to the licensing terms of the authors. Moreover, if you will be making use of this particular implementation[3], please acknowledge it.

	GitHub	Ref
Original version based on Caffe	https://github.com/s9xie/hed	[1]
Another reimplementation based on Caffe	https://github.com/zeakey/hed
Original reimplementation based on PyTorch	https://github.com/sniklaus/pytorch-hed	[2]

Usage

First, you have to install the package (stable) with

pip install pytorch-hed

or, for the current (unstable) version

pip install git+https://github.com/Davidelanz/pytorch-hed.git

Usage:

import torchHED
   
# process a single image file 
torchHED.process_file("./images/sample.png", "./images/sample_processed.png")

# process all images in a folder
torchHED.process_folder("./input_folder", "./output_folder")

# process a PIL.Image loaded in memory and return a new PIL.Image
# img = PIL.Image.open("./images/sample.png")
img_hed = torchHED.process_img(img)

Results

Input	Original Caffe Implementation [1]	pytorch-hed [3]

References

[1]  @inproceedings{Xie_ICCV_2015,
         author = {Saining Xie and Zhuowen Tu},
         title = {Holistically-Nested Edge Detection},
         booktitle = {IEEE International Conference on Computer Vision},
         year = {2015}
     }

[2]  @misc{pytorch-hed,
         author = {Simon Niklaus},
         title = {A Reimplementation of {HED} Using {PyTorch}},
         year = {2018},
         howpublished = {\url{https://github.com/sniklaus/pytorch-hed}}
    }

[3]  @misc{pytorch-hed-2,
         author = {Davide Lanza},
         title = {The {pytorch-hed} Python Package},
         year = {2020},
         howpublished = {\url{https://github.com/Davidelanz/pytorch-hed}}
    }