OpenPifPaf: Composite Fields for Semantic Keypoint Detection and Spatio-Temporal Association

3 Mar 2021  ·  Sven Kreiss, Lorenzo Bertoni, Alexandre Alahi ·

Many image-based perception tasks can be formulated as detecting, associating and tracking semantic keypoints, e.g., human body pose estimation and tracking. In this work, we present a general framework that jointly detects and forms spatio-temporal keypoint associations in a single stage, making this the first real-time pose detection and tracking algorithm. We present a generic neural network architecture that uses Composite Fields to detect and construct a spatio-temporal pose which is a single, connected graph whose nodes are the semantic keypoints (e.g., a person's body joints) in multiple frames. For the temporal associations, we introduce the Temporal Composite Association Field (TCAF) which requires an extended network architecture and training method beyond previous Composite Fields. Our experiments show competitive accuracy while being an order of magnitude faster on multiple publicly available datasets such as COCO, CrowdPose and the PoseTrack 2017 and 2018 datasets. We also show that our method generalizes to any class of semantic keypoints such as car and animal parts to provide a holistic perception framework that is well suited for urban mobility such as self-driving cars and delivery robots.

PDF Abstract

Results from the Paper


Task Dataset Model Metric Name Metric Value Global Rank Result Benchmark
Car Pose Estimation ApolloCar3D OpenPifPaf Detection Rate 86.1 # 2
Keypoint Detection COCO test-dev OpenPifPaf APL 76.8 # 9
APM 67.1 # 8
AP 70.9 # 2
Pose Estimation CrowdPose OpenPifPaf AP 70.5 # 7
AP50 89.1 # 2
AP75 76.1 # 2
AP Hard 63.8 # 3
AP Easy 78.4 # 2
AP Medium 72.1 # 1
Multi-Person Pose Estimation MS COCO OpenPifPaf AP 0.709 # 7
Validation AP 71.0 # 4
Test AP 70.9 # 5

Methods