SMPR: Single-Stage Multi-Person Pose Regression

TL;DR

SMPR introduces a novel single-stage multi-person pose regression method that predicts instance-aware keypoints densely, improving performance and efficiency over existing single-stage and bottom-up approaches on the COCO benchmark.

Contribution

The paper proposes a new dense prediction framework for multi-person pose estimation that enhances positive hypothesis selection and pose scoring, achieving state-of-the-art results.

Findings

Outperforms existing single-stage methods with 70.2 AP on COCO.

Competitive with latest bottom-up approaches.

Effective pose scoring strategy improves NMS performance.

Abstract

Existing multi-person pose estimators can be roughly divided into two-stage approaches (top-down and bottom-up approaches) and one-stage approaches. The two-stage methods either suffer high computational redundancy for additional person detectors or group keypoints heuristically after predicting all the instance-free keypoints. The recently proposed single-stage methods do not rely on the above two extra stages but have lower performance than the latest bottom-up approaches. In this work, a novel single-stage multi-person pose regression, termed SMPR, is presented. It follows the paradigm of dense prediction and predicts instance-aware keypoints from every location. Besides feature aggregation, we propose better strategies to define positive pose hypotheses for training which all play an important role in dense pose estimation. The network also learns the scores of estimated poses. The pose scoring strategy further improves the pose estimation performance by prioritizing superior poses during non-maximum suppression (NMS). We show that our method not only outperforms existing single-stage methods and but also be competitive with the latest bottom-up methods, with 70.2 AP and 77.5 AP75 on the COCO test-dev pose benchmark. Code is available at https://github.com/cmdi-dlut/SMPR.