Bootstrapping Human Optical Flow and Pose
This work addresses the challenge of enhancing human motion analysis in computer vision, which is incremental by building on existing methods for optical flow and pose estimation.
The paper tackles the problem of improving human optical flow and pose estimation in videos by proposing a bootstrapping framework that jointly optimizes both tasks, resulting in state-of-the-art performance on datasets like Human 3.6M and 3D Poses in the Wild.
We propose a bootstrapping framework to enhance human optical flow and pose. We show that, for videos involving humans in scenes, we can improve both the optical flow and the pose estimation quality of humans by considering the two tasks at the same time. We enhance optical flow estimates by fine-tuning them to fit the human pose estimates and vice versa. In more detail, we optimize the pose and optical flow networks to, at inference time, agree with each other. We show that this results in state-of-the-art results on the Human 3.6M and 3D Poses in the Wild datasets, as well as a human-related subset of the Sintel dataset, both in terms of pose estimation accuracy and the optical flow accuracy at human joint locations. Code available at https://github.com/ubc-vision/bootstrapping-human-optical-flow-and-pose