On the power of data augmentation for head pose estimation

Document detail

ID

oai:arXiv.org:2407.05357

Topic

Computer Science - Computer Vision...

Author

Welter, Michael

Year

2024

listing date

10/23/2024

Keywords

estimation

Metrics

Abstract

Deep learning has been impressively successful in the last decade in predicting human head poses from monocular images.

However, for in-the-wild inputs the research community relies predominantly on a single training set, 300W-LP, of semisynthetic nature without many alternatives.

This paper focuses on gradual extension and improvement of the data to explore the performance achievable with augmentation and synthesis strategies further.

Modeling-wise a novel multitask head/loss design which includes uncertainty estimation is proposed.

Overall, the thus obtained models are small, efficient, suitable for full 6 DoF pose estimation, and exhibit very competitive accuracy.

;Comment: CVPR version.

Added evaluation on BIWI.

Plenty of writing changes

Welter, Michael, 2024, On the power of data augmentation for head pose estimation