Towards Accurate Multi-person Pose Estimation in the Wild

We propose a method for multi-person detection and 2-D pose estimation that achieves state-of-art results on the challenging COCO keypoints task. It is a simple, yet powerful, top-down approach consisting of two stages... (read more)

PDF Abstract CVPR 2017 PDF CVPR 2017 Abstract
No code implementations yet. Submit your code now
TASK DATASET MODEL METRIC NAME METRIC VALUE GLOBAL RANK RESULT BENCHMARK
Multi-Person Pose Estimation COCO G-RMI* AP 0.685 # 6
Multi-Person Pose Estimation COCO G-RMI AP 0.649 # 8
Keypoint Detection COCO test-challenge G-RMI* AR 75.1 # 5
ARM 69.7 # 5
AP 69.1 # 4
AP50 85.9 # 5
AP75 75.2 # 4
APL 82.4 # 5
AR50 90.7 # 5
AR75 80.7 # 5
ARL 74.5 # 5
Keypoint Detection COCO test-dev G-RMI APL 70.0 # 14
APM 62.3 # 10
AP50 85.5 # 11
AP75 71.3 # 9
AR 69.7 # 9
AR50 88.7 # 6
AR75 75.5 # 6
ARL 77.1 # 6
ARM 64.4 # 6
Pose Estimation COCO test-dev G-RMI AP 64.9 # 13
AP50 85.5 # 15
AP75 71.3 # 13
APL 70.0 # 14
AR 69.7 # 12
Multi-Person Pose Estimation COCO test-dev G-RMI AP 64.9 # 10
APL 70.0 # 7
APM 62.3 # 7
AP50 85.5 # 6
AP75 71.3 # 6

Methods used in the Paper


METHOD TYPE
Average Pooling
Pooling Operations
Global Average Pooling
Pooling Operations
1x1 Convolution
Convolutions
ReLU
Activation Functions
Batch Normalization
Normalization
Bottleneck Residual Block
Skip Connection Blocks
Max Pooling
Pooling Operations
Kaiming Initialization
Initialization
Residual Connection
Skip Connections
Convolution
Convolutions
Residual Block
Skip Connection Blocks
ResNet
Convolutional Neural Networks