Kinematic-aware Hierarchical Attention Network for Human Pose Estimation in Videos
This work improves pose estimation accuracy and smoothness in videos, which is important for applications like motion analysis and human-computer interaction, though it appears incremental as it builds on existing estimators with novel temporal feature integration.
The paper tackles the problem of video-based human pose estimation by addressing jitter, insufficient temporal comprehension, and occlusion issues, resulting in a model that demonstrates effectiveness across 2D pose estimation, 3D pose estimation, body mesh recovery, and sparsely annotated multi-human pose estimation tasks.
Previous video-based human pose estimation methods have shown promising results by leveraging aggregated features of consecutive frames. However, most approaches compromise accuracy to mitigate jitter or do not sufficiently comprehend the temporal aspects of human motion. Furthermore, occlusion increases uncertainty between consecutive frames, which results in unsmooth results. To address these issues, we design an architecture that exploits the keypoint kinematic features with the following components. First, we effectively capture the temporal features by leveraging individual keypoint's velocity and acceleration. Second, the proposed hierarchical transformer encoder aggregates spatio-temporal dependencies and refines the 2D or 3D input pose estimated from existing estimators. Finally, we provide an online cross-supervision between the refined input pose generated from the encoder and the final pose from our decoder to enable joint optimization. We demonstrate comprehensive results and validate the effectiveness of our model in various tasks: 2D pose estimation, 3D pose estimation, body mesh recovery, and sparsely annotated multi-human pose estimation. Our code is available at https://github.com/KyungMinJin/HANet.