xindongzhang/deep-head-pose

:fire::fire: Deep Learning Head Pose Estimation using PyTorch.

Hopenet

Hopenet is an accurate and easy to use head pose estimation network. Models have been trained on the 300W-LP dataset and have been tested on real data with good qualitative performance.

For details about the method and quantitative results please check the paper.

new GoT trailer example video

new Conan-Cruise-Car example video

To use please install PyTorch and OpenCV (for video) - I believe that's all you need apart from usual libraries such as numpy. You need a GPU to run Hopenet (for now).

To test on a video using dlib face detections (center of head will be jumpy):

python code/test_on_video_dlib.py --snapshot PATH_OF_SNAPSHOT --face_model PATH_OF_DLIB_MODEL --video PATH_OF_VIDEO --output_string STRING_TO_APPEND_TO_OUTPUT --n_frames N_OF_FRAMES_TO_PROCESS --fps FPS_OF_SOURCE_VIDEO

To test on a video using your own face detections (we recommend using dockerface, center of head will be smoother):

python code/test_on_video_dockerface.py --snapshot PATH_OF_SNAPSHOT --video PATH_OF_VIDEO --bboxes FACE_BOUNDING_BOX_ANNOTATIONS --output_string STRING_TO_APPEND_TO_OUTPUT --n_frames N_OF_FRAMES_TO_PROCESS --fps FPS_OF_SOURCE_VIDEO

Face bounding box annotations should be in Dockerface format (n_frame x_min y_min x_max y_max confidence).

Pre-trained models:

300W-LP, alpha 1

300W-LP, alpha 2

300W-LP, alpha 1, robust to image quality

For more information on what alpha stands for please read the paper. First two models are for validating paper results, if used on real data we suggest using the last model as it is more robust to image quality and blur and gives good results on video.

Please keep in mind that testing instructions to reproduce the paper results will be added very soon.

This work is still in progress - we are obtaining better results and will also be updating this README with instructions. Please open an issue if you have an problem.

Some things that will be added:

• Test script for images
• Docker image
• Instructions for all scripts
• Better and better models
• Videos and example images!

If you find Hopenet useful in your research please consider citing:

@article{DBLP:journals/corr/abs-1710-00925,
  author    = {Nataniel Ruiz and
               Eunji Chong and
               James M. Rehg},
  title     = {Fine-Grained Head Pose Estimation Without Keypoints},
  journal   = {CoRR},
  volume    = {abs/1710.00925},
  year      = {2017},
  url       = {http://arxiv.org/abs/1710.00925},
  archivePrefix = {arXiv},
  eprint    = {1710.00925},
  timestamp = {Wed, 01 Nov 2017 19:05:43 +0100},
  biburl    = {http://dblp.org/rec/bib/journals/corr/abs-1710-00925},
  bibsource = {dblp computer science bibliography, http://dblp.org}
}

Nataniel Ruiz, Eunji Chong, James M. Rehg

Georgia Institute of Technology