SRN

A Tensorflow implementation of the paper: Reasoning structural relation for occlusion-robust facial landmark localization.

Installation Instructions

Menpo 0.8.1

Menpodetect 0.5.0

Menpo fit 0.5.0

we use the Menpo project in various ways throughout the implementation.

Please look at the installation instructions at:

http://www.menpo.org/installation/

TensorFlow 1.10.1

Pretrained models

The pre-training model is coming soon.

Training a model

Currently the TensorFlow implementation does not contain tracking model we did in the submitted paper, but this will be updated shortly.

    # Activate the conda environment.
    source activate environment-name
    
    # Start training
    python train.py --datasets='databases/lfpw/trainset/*.png:databases/afw/*.jpg:databases/helen/trainset/*.jpg'
    
    # Track the train process and evaluate the current checkpoint against the validation set
    python eval.py --dataset_path="./databases/ibug/*.jpg" --num_examples=135 --eval_dir=ckpt/eval_ibug  --device='/cpu:0' --checkpoint_dir=$PWD/ckpt/train
    
    python eval.py --dataset_path="./databases/lfpw/testset/*.png" --num_examples=300 --eval_dir=ckpt/eval_lfpw  --device='/cpu:0' --checkpoint_dir=$PWD/ckpt/train
    
    python eval.py --dataset_path="./databases/helen/testset/*.jpg" --num_examples=330 --eval_dir=ckpt/eval_helen  --device='/cpu:0' --checkpoint_dir=$PWD/ckpt/train
    
    # Run tensorboard to visualise the results
    tensorboard --logdir==$PWD/ckpt

The implementation of some functions refers to the MDM project (https://github.com/trigeorgis/mdm).