AdaptivePose++: A Powerful Single-Stage Network for Multi-Person Pose Regression

TL;DR

AdaptivePose++ introduces a single-stage network for multi-person pose estimation that encodes human parts as adaptive points, enabling accurate 2D and 3D pose estimation with reduced complexity and high efficiency.

Contribution

The paper proposes a novel adaptive body representation and a compact single-stage network for multi-person pose regression, eliminating the need for complex post-processing.

Findings

Achieves state-of-the-art accuracy on MS COCO and CrowdPose datasets.

Operates with high speed and efficiency in a single forward pass.

Demonstrates strong generalization to 3D pose estimation tasks.

Abstract

Multi-person pose estimation generally follows top-down and bottom-up paradigms. Both of them use an extra stage ($\boldsymbol{e.g.,}$ 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. To address the above issue, we propose to represent the human parts as adaptive points and introduce a fine-grained body representation method. The novel body representation is able to sufficiently encode the diverse pose information and effectively model the relationship between the human instance and corresponding keypoints in a single-forward pass. With the proposed body representation, we further deliver a compact single-stage multi-person pose regression network, termed as AdaptivePose. During inference, our proposed network only needs a single-step decode operation to form the multi-person pose without complex post-processes and refinements. We employ AdaptivePose for both 2D/3D multi-person pose estimation tasks to verify the effectiveness of AdaptivePose. Without any bells and whistles, we achieve the most competitive performance on MS COCO and CrowdPose in terms of accuracy and speed. Furthermore, the outstanding performance on MuCo-3DHP and MuPoTS-3D further demonstrates the effectiveness and generalizability on 3D scenes. Code is available at https://github.com/buptxyb666/AdaptivePose.