sellGuided Research sellDFKI sellAugmented Vision sell3D Human Pose Estimation

Monocular 3D Human Pose Estimation Using Depth Augmentation (Guided Research)

Jeremias Krauss , Muhammad Saif Ullah Khan

Abstract

This guided research investigates how depth-aware augmentation can improve monocular 3D human pose estimation. The work studies synthetic and pseudo-depth cues to reduce geometric ambiguity and improve 3D keypoint consistency.

Topic

Study monocular 3D human pose estimation with explicit depth augmentation to improve robustness and geometric consistency under viewpoint and occlusion changes.

Tasks

1. Prepare a monocular 3D pose baseline and evaluation protocol.
2. Implement depth-augmentation strategies for training.
3. Evaluate gains in 3D keypoint accuracy and temporal consistency.

Related Literature

• [1] A Simple Yet Effective Baseline for 3D Human Pose Estimation
• [2] VideoPose3D: 3D Human Pose Estimation in Video With Temporal Convolutions
• [3] Monocular 3D Human Pose Estimation in the Wild

Maintained by saifkhichi96 on GitHub.

The website is distributed under different open-source licenses. For more details, see the notice at the bottom of the page.