Convolutional Pose Machines
Pose Machines provide a sequential prediction framework for learning rich implicit spatial models. In this work we show a systematic design for how convolutional networks can be incorporated into the pose machine framework for learning image features and image-dependent spatial models for the task of pose estimation. The contribution of this paper is to implicitly model long-range dependencies between variables in structured prediction tasks such as articulated pose estimation. We achieve this by designing a sequential architecture composed of convolutional networks that directly operate on belief maps from previous stages, producing increasingly refined estimates for part locations, without the need for explicit graphical model-style inference. Our approach addresses the characteristic difficulty of vanishing gradients during training by providing a natural learning objective function that enforces intermediate supervision, thereby replenishing back-propagated gradients and conditioning the learning procedure. We demonstrate state-of-the-art performance and outperform competing methods on standard benchmarks including the MPII, LSP, and FLIC datasets.
PDF Abstract CVPR 2016 PDF CVPR 2016 AbstractCode
Datasets
MPII
JHMDB
LSP
MPII Human Pose
TotalCapture
FLIC
RSSCN7
ApolloCar3D
| Task | Dataset | Model | Metric Name | Metric Value | Global Rank | Benchmark |
|---|---|---|---|---|---|---|
| Car Pose Estimation | ApolloCar3D | CPM | Detection Rate | 75.4 | # 3 | |
| Pose Estimation | FLIC Elbows | Convolutional Pose Machines | PCK@0.2 | 97.59% | # 2 | |
| Pose Estimation | FLIC Wrists | Convolutional Pose Machines | PCK@0.2 | 95.03% | # 2 | |
| Classification | RSSCN7 | CPM | 1:1 Accuracy | 50 | # 2 |
