Loading
Cookies help us deliver our services. By using our services, you agree to our use of cookies. Learn more

Dark Skies

Tensorflow solution for the Dark Skies Challenge

I use Inception v3 architecture as described in “Rethinking the Inception Architecture for Computer Vision” by Christian Szegedy, Vincent Vanhoucke, Sergey Ioffe, Jonathon Shlens, Zbigniew Wojna. The original paper can be access here: https://arxiv.org/abs/1512.00567.


inception_v3_architecture.png
Schematic diagram of Inception V3

To speed up the training process, I use the model that is trained on ImageNet 1k dataset. It is released by Google. These are steps to download:

# set $DK_ROOT to the folder that is cloned from darkskies-challenge github repo 
# (https://github.com/LiberiFatali/darkskies-challenge)
# for example: 
DK_ROOT=/home/username/darkskies-challenge
# location of where to place the Inception v3 model
DATA_DIR=$DK_ROOT/inception-v3-model
cd ${DATA_DIR}

# download the Inception v3 model
curl -O http://download.tensorflow.org/models/image/imagenet/inception-v3-2016-03-01.tar.gz
# extract
tar xzf inception-v3-2016-03-01.tar.gz

# this will create a directory called inception-v3 which contains the following files.
> ls inception-v3
README.txt
checkpoint
model.ckpt-157585