Salience Object Detection Using python + tensorflow This is tensorflow implementation for cvpr2017 paper "Deeply Supervised Salient Object Detection with Short Connections".
If it is useful for u, pls give me a star! THX
Cuz i trained this model on my company's data,so to be honest i dont know this model still reach the performance mentioned in paper or not.
This code can be used to train, but the data is owned by my company.I'll try my best to provide code and model that can do inference.
My Chinese blog about the implementation of this paper http://blog.csdn.net/gbyy42299/article/details/79427457
I trained this model on gpu Tesla P100(1200 images, 10h)
u need download vgg16.npy as pre-trained model.
1.Just put ur data under the folder named 'dataset'.
2.run csv_generater.py to get the csv files for ur training.
3.run train.py for training ur own model.
4.run test.py for testing new photo.
I added model scripy fot u guys for training on ur own dataset.
I fixed some bug, like loss problem etc, now u can use it.
U can find another salience obeject detection in this url: https://github.com/gbyy422990/salience_object-detection-non-local