Deep Kinematic Pose Regression
This addresses the challenge of maintaining structural constraints in pose estimation for applications like robotics or human motion analysis, though it is incremental by building on existing kinematic modeling.
The authors tackled the problem of articulated object pose estimation by embedding a kinematic model into a deep neural network to ensure geometric validity, achieving state-of-the-art results on the Human3.6M dataset.
Learning articulated object pose is inherently difficult because the pose is high dimensional but has many structural constraints. Most existing work do not model such constraints and does not guarantee the geometric validity of their pose estimation, therefore requiring a post-processing to recover the correct geometry if desired, which is cumbersome and sub-optimal. In this work, we propose to directly embed a kinematic object model into the deep neutral network learning for general articulated object pose estimation. The kinematic function is defined on the appropriately parameterized object motion variables. It is differentiable and can be used in the gradient descent based optimization in network training. The prior knowledge on the object geometric model is fully exploited and the structure is guaranteed to be valid. We show convincing experiment results on a toy example and the 3D human pose estimation problem. For the latter we achieve state-of-the-art result on Human3.6M dataset.