Faster and better multi-person pose estimation by using a single stage with AdaptivePose
Faster and better multi-person pose estimation by using a single stage with AdaptivePose
AdaptivePose++: A Powerful Single-Stage Network for Multi-Person Pose Regression
arXiv paper abstract https://arxiv.org/abs/2210.04014v1
arXiv PDF paper https://arxiv.org/pdf/2210.04014v1.pdf
Multi-person pose estimation generally follows top-down and bottom-up paradigms.
Both of them use an extra stage (e.g., human detection in top-down paradigm or grouping process in bottom-up paradigm) to build the relationship between the human instance and corresponding keypoints, thus leading to the high computation cost and redundant two-stage pipeline.
... propose to represent the human parts as adaptive points and introduce a fine-grained body representation method.
... deliver a compact single-stage multi-person pose regression network, termed as AdaptivePose.
During inference ... only needs a single-step decode operation to form the multi-person pose without complex post-processes and refinements.
... employ AdaptivePose for both 2D/3D multi-person pose estimation ... achieve the most competitive performance on MS COCO and CrowdPose in terms of accuracy and speed ...
Please like and share this post if you enjoyed it using the buttons at the bottom!
Stay up to date. Subscribe to my posts https://morrislee1234.wixsite.com/website/contact
Web site with my other posts by category https://morrislee1234.wixsite.com/website
Comments