"Deep Head Pose Estimation Using Synthetic Images and Partial Adversarial Domain Adaption for Continuous Label Spaces" ICCV 2019

(pdf)

Update Jan. 2023

A python version of Biwi+, our models and evaluation code can be found here GitHub
The matlab version has been updated too.

The HeadPose+ Dataset

The dataset includes the datasets:
- SynHead++ (for partial domain adaptation, source dataset)
- SynBiwi+ (for domain adaptation, source dataset)
- Biwi+ (for head pose evaluation, target dataset)

We provide consistent/comparable head pose annotations and bounding boxes for all three datasets. To create the dataset you have to download Biwi and SynHead and the provided code and labels. Please follow the readme.txt instructions.

HeadPose+ (zip file)

Code and labels are restricted for research purposes only. By using the datasets/code, you have to acknowledge the effort by citing the corresponding papers (see readme.txt).
If you have any questions please contact me