A small example package
Project description
tensorflow-object-contour-detection
This is a tensorflow implimentation of Object Contour Detection with a Fully Convolutional Encoder-Decoder Network (https://arxiv.org/pdf/1603.04530.pdf) .
REQUIREMENTS :
pip install requirements.txt
Label Preparation :
To prepare the labels for contour detection from PASCAL Dataset , run create_lables.py and edit the file to add the path of the labels and new labels to be generated . Use this path for labels during training.
TRAINING :
python train.py \
--max_to_keep=50 \
--Epochs=100 \
--momentum=0.9 \
--learning_rate=.0000001 \
--train_crop_size=480 \
--clip_by_value=1.0 \
--train_text = ${path to text file} \
--log_dir = ${path to where logs will be saved} \
--tf_initial_checkpoint=${PATH_TO_CHECKPOINT} \
--label_dir = ${path to label directory} \
--image_dir = ${path to image directory}
EVALUATION :
python eval.py \
--checkpoint=${path to checkpoint to be evaluated} \
--save_preds=${path to folder where predictions will be saved} \
--image_dir = ${path to image directory} \
--eval_crop_size=480 \
--eval_text = ${path to eval text file}
Results :
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
Close
Hashes for hubisModel-pkg-yjko-0.0.1.tar.gz
Algorithm | Hash digest | |
---|---|---|
SHA256 | 6f427c5bdd3a6e3faa65f92ee605952a2ef5ca33162f33a447dd8c0e7327b73d |
|
MD5 | 365a9625d41e093fec33f05cc783f2e6 |
|
BLAKE2b-256 | 5c483fb6b900359f5500ab76a14abd83bd60c13c6390e891d14a981f43c7304f |
Close
Hashes for hubisModel_pkg_yjko-0.0.1-py3-none-any.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | c733841449234e635114dbdaca424838a6cd4ef431724f761cce9674013aff15 |
|
MD5 | 98e37c7618d31c6d497ddad308de4e3c |
|
BLAKE2b-256 | 5918d8f3a8b58e59d7b2189dd0711659a2c323293f1255ad587bb999a8a07ebf |