TASK	DATASET	MODEL	METRIC NAME	METRIC VALUE	GLOBAL RANK
Pose Estimation	AIC	SimpleBaseline (ResNet-101)	AP	29.4	# 9
Pose Estimation	AIC	SimpleBaseline (ResNet-101)	AP50	73.6	# 5
Pose Estimation	AIC	SimpleBaseline (ResNet-101)	AP75	17.4	# 6
Pose Estimation	AIC	SimpleBaseline (ResNet-101)	AR	33.7	# 6
Pose Estimation	AIC	SimpleBaseline (ResNet-101)	AR50	76.3	# 6
Pose Estimation	AIC	SimpleBaseline (ResNet-152)	AP	29.9	# 8
Pose Estimation	AIC	SimpleBaseline (ResNet-152)	AP50	73.8	# 4
Pose Estimation	AIC	SimpleBaseline (ResNet-152)	AP75	18.3	# 5
Pose Estimation	AIC	SimpleBaseline (ResNet-152)	AR	34.3	# 5
Pose Estimation	AIC	SimpleBaseline (ResNet-152)	AR50	76.9	# 5
Pose Estimation	AIC	SimpleBaseline (ResNet-50)	AP	28.0	# 10
Pose Estimation	AIC	SimpleBaseline (ResNet-50)	AP50	71.6	# 6
Pose Estimation	AIC	SimpleBaseline (ResNet-50)	AP75	15.8	# 7
Pose Estimation	AIC	SimpleBaseline (ResNet-50)	AR	32.1	# 7
Pose Estimation	AIC	SimpleBaseline (ResNet-50)	AR50	74.1	# 7
Keypoint Detection	COCO test-challenge	Simple Base+*	AR	80.5	# 3
Keypoint Detection	COCO test-challenge	Simple Base+*	ARM	75.3	# 3
Keypoint Detection	COCO test-challenge	Simple Base+*	AP	74.5	# 3
Keypoint Detection	COCO test-challenge	Simple Base+*	AP50	90.9	# 3
Keypoint Detection	COCO test-challenge	Simple Base+*	AP75	80.8	# 3
Keypoint Detection	COCO test-challenge	Simple Base+*	APL	87.5	# 2
Keypoint Detection	COCO test-challenge	Simple Base+*	AR50	95.1	# 3
Keypoint Detection	COCO test-challenge	Simple Base+*	AR75	86.3	# 3
Keypoint Detection	COCO test-challenge	Simple Base+*	ARL	82.9	# 3
Keypoint Detection	COCO test-dev	Simple Base	APL	80.0	# 6
Keypoint Detection	COCO test-dev	Simple Base	APM	70.3	# 5
Keypoint Detection	COCO test-dev	Simple Base	AP50	91.9	# 5
Keypoint Detection	COCO test-dev	Simple Base	AP75	81.1	# 5
Keypoint Detection	COCO test-dev	Simple Base	AR	79.0	# 5
Pose Estimation	COCO test-dev	Flow-based (ResNet-152)	AP	73.7	# 24
Pose Estimation	COCO test-dev	Flow-based (ResNet-152)	AP50	91.9	# 18
Pose Estimation	COCO test-dev	Flow-based (ResNet-152)	AP75	81.1	# 21
Pose Estimation	COCO test-dev	Flow-based (ResNet-152)	APL	80	# 16
Pose Estimation	COCO test-dev	Flow-based (ResNet-152)	APM	70.3	# 21
Pose Estimation	COCO test-dev	Flow-based (ResNet-152)	AR	79	# 20
Keypoint Detection	COCO test-dev	Simple Base+*	APL	82.7	# 2
Keypoint Detection	COCO test-dev	Simple Base+*	APM	73.0	# 2
Keypoint Detection	COCO test-dev	Simple Base+*	AP50	92.4	# 4
Keypoint Detection	COCO test-dev	Simple Base+*	AP75	84.0	# 2
Keypoint Detection	COCO test-dev	Simple Base+*	AR	81.5	# 3
Keypoint Detection	COCO test-dev	Simple Base+*	AR50	95.8	# 2
Keypoint Detection	COCO test-dev	Simple Base+*	AR75	88.2	# 1
Keypoint Detection	COCO test-dev	Simple Base+*	ARL	87.2	# 1
Keypoint Detection	COCO test-dev	Simple Base+*	ARM	77.4	# 2
Multi-Person Pose Estimation	CrowdPose	Simple baseline	mAP @0.5:0.95	60.8	# 18
Multi-Person Pose Estimation	CrowdPose	Simple baseline	AP Easy	71.4	# 14
Multi-Person Pose Estimation	CrowdPose	Simple baseline	AP Medium	61.2	# 17
Multi-Person Pose Estimation	CrowdPose	Simple baseline	AP Hard	51.2	# 15
2D Human Pose Estimation	JHMDB (2D poses only)	SimplePose	PCK	94.4	# 2
Keypoint Detection	MS COCO	ResNet-50	Validation AP	72.2	# 10
Pose Estimation	OCHuman	ResNet-152	Test AP	33.3	# 8
Pose Estimation	OCHuman	ResNet-152	Validation AP	41.0	# 7
Keypoint Detection	OCHuman	ResNet-50	Test AP	29.5	# 7
Keypoint Detection	OCHuman	ResNet-50	Validation AP	32.1	# 7
Keypoint Detection	OCHuman	ResNet-152	Test AP	33.3	# 4
Keypoint Detection	OCHuman	ResNet-152	Validation AP	41.0	# 3
2D Human Pose Estimation	OCHuman	ResNet-50	Test AP	30.4	# 8
2D Human Pose Estimation	OCHuman	ResNet-50	Validation AP	37.8	# 6
2D Human Pose Estimation	OCHuman	ResNet-152	Test AP	33.3	# 5
2D Human Pose Estimation	OCHuman	ResNet-152	Validation AP	41.0	# 3
Multi-Person Pose Estimation	OCHuman	SimplePose	Validation AP	24.1	# 6
Multi-Person Pose Estimation	OCHuman	SimplePose	AP50	37.4	# 7
Multi-Person Pose Estimation	OCHuman	SimplePose	AP75	26.8	# 7
Pose Estimation	OCHuman	ResNet-50	Test AP	29.5	# 11
Pose Estimation	OCHuman	ResNet-50	Validation AP	32.1	# 11
Pose Tracking	PoseTrack2017	MSRA (FlowTrack)	MOTA	57.81	# 5
Pose Tracking	PoseTrack2017	MSRA (FlowTrack)	mAP	74.57	# 2
Pose Tracking	PoseTrack2018	MSRA	MOTA	61.37	# 4
Pose Tracking	PoseTrack2018	MSRA	mAP	74.03	# 1

Badge	Markdown
	`[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/simple-baselines-for-human-pose-estimation/keypoint-detection-on-coco-test-dev)](https://paperswithcode.com/sota/keypoint-detection-on-coco-test-dev?p=simple-baselines-for-human-pose-estimation)`
	`[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/simple-baselines-for-human-pose-estimation/2d-human-pose-estimation-on-jhmdb-2d-poses)](https://paperswithcode.com/sota/2d-human-pose-estimation-on-jhmdb-2d-poses?p=simple-baselines-for-human-pose-estimation)`
	`[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/simple-baselines-for-human-pose-estimation/keypoint-detection-on-coco-test-challenge)](https://paperswithcode.com/sota/keypoint-detection-on-coco-test-challenge?p=simple-baselines-for-human-pose-estimation)`
	`[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/simple-baselines-for-human-pose-estimation/keypoint-detection-on-ochuman)](https://paperswithcode.com/sota/keypoint-detection-on-ochuman?p=simple-baselines-for-human-pose-estimation)`
	`[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/simple-baselines-for-human-pose-estimation/pose-tracking-on-posetrack2018)](https://paperswithcode.com/sota/pose-tracking-on-posetrack2018?p=simple-baselines-for-human-pose-estimation)`
	`[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/simple-baselines-for-human-pose-estimation/2d-human-pose-estimation-on-ochuman)](https://paperswithcode.com/sota/2d-human-pose-estimation-on-ochuman?p=simple-baselines-for-human-pose-estimation)`
	`[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/simple-baselines-for-human-pose-estimation/pose-tracking-on-posetrack2017)](https://paperswithcode.com/sota/pose-tracking-on-posetrack2017?p=simple-baselines-for-human-pose-estimation)`
	`[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/simple-baselines-for-human-pose-estimation/multi-person-pose-estimation-on-ochuman)](https://paperswithcode.com/sota/multi-person-pose-estimation-on-ochuman?p=simple-baselines-for-human-pose-estimation)`
	`[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/simple-baselines-for-human-pose-estimation/pose-estimation-on-aic)](https://paperswithcode.com/sota/pose-estimation-on-aic?p=simple-baselines-for-human-pose-estimation)`
	`[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/simple-baselines-for-human-pose-estimation/pose-estimation-on-ochuman)](https://paperswithcode.com/sota/pose-estimation-on-ochuman?p=simple-baselines-for-human-pose-estimation)`
	`[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/simple-baselines-for-human-pose-estimation/keypoint-detection-on-coco)](https://paperswithcode.com/sota/keypoint-detection-on-coco?p=simple-baselines-for-human-pose-estimation)`
	`[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/simple-baselines-for-human-pose-estimation/multi-person-pose-estimation-on-crowdpose)](https://paperswithcode.com/sota/multi-person-pose-estimation-on-crowdpose?p=simple-baselines-for-human-pose-estimation)`
	`[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/simple-baselines-for-human-pose-estimation/pose-estimation-on-coco-test-dev)](https://paperswithcode.com/sota/pose-estimation-on-coco-test-dev?p=simple-baselines-for-human-pose-estimation)`

Simple Baselines for Human Pose Estimation and Tracking

ECCV 2018 · Bin Xiao, Haiping Wu, Yichen Wei ·

There has been significant progress on pose estimation and increasing interests on pose tracking in recent years. At the same time, the overall algorithm and system complexity increases as well, making the algorithm analysis and comparison more difficult. This work provides simple and effective baseline methods. They are helpful for inspiring and evaluating new ideas for the field. State-of-the-art results are achieved on challenging benchmarks. The code will be available at https://github.com/leoxiaobin/pose.pytorch.

PDF Abstract ECCV 2018 PDF ECCV 2018 Abstract

Code

Add Remove Mark official

leoxiaobin/pose.pytorch official

PaddlePaddle/PaddleDetection

12,048

open-mmlab/mmpose

4,986

osmr/imgclsmob

2,917

Microsoft/human-pose-estimation.pyt…

2,905

See all 27 implementations

Tasks

Add Remove

2D Human Pose Estimation

Keypoint Detection

Multi-Person Pose Estimation

Pose Estimation

Pose Tracking

Datasets

MS COCO

MPII

JHMDB

PoseTrack

CrowdPose

OCHuman AIC

Results from the Paper

Edit

Ranked #2 on 2D Human Pose Estimation on JHMDB (2D poses only)

Get a GitHub badge

Task	Dataset	Model	Metric Name	Metric Value	Global Rank	Benchmark
Pose Estimation	AIC	SimpleBaseline (ResNet-101)	AP	29.4	# 9	Compare
			AP50	73.6	# 5	Compare
			AP75	17.4	# 6	Compare
			AR	33.7	# 6	Compare
			AR50	76.3	# 6	Compare
Keypoint Detection	COCO test-challenge	Simple Base+*	AR	80.5	# 3	Compare
			ARM	75.3	# 3	Compare
			AP	74.5	# 3	Compare
			AP50	90.9	# 3	Compare
			AP75	80.8	# 3	Compare
			APL	87.5	# 2	Compare
			AR50	95.1	# 3	Compare
			AR75	86.3	# 3	Compare
			ARL	82.9	# 3	Compare
Keypoint Detection	COCO test-dev	Simple Base	APL	80.0	# 6	Compare
			APM	70.3	# 5	Compare
			AP50	91.9	# 5	Compare
			AP75	81.1	# 5	Compare
			AR	79.0	# 5	Compare
Pose Estimation	COCO test-dev	Flow-based (ResNet-152)	AP	73.7	# 24	Compare
			AP50	91.9	# 18	Compare
			AP75	81.1	# 21	Compare
			APL	80	# 16	Compare
			APM	70.3	# 21	Compare
			AR	79	# 20	Compare
Keypoint Detection	COCO test-dev	Simple Base+*	APL	82.7	# 2	Compare
			APM	73.0	# 2	Compare
			AP50	92.4	# 4	Compare
			AP75	84.0	# 2	Compare
			AR	81.5	# 3	Compare
			AR50	95.8	# 2	Compare
			AR75	88.2	# 1	Compare
			ARL	87.2	# 1	Compare
			ARM	77.4	# 2	Compare
Multi-Person Pose Estimation	CrowdPose	Simple baseline	mAP @0.5:0.95	60.8	# 18	Compare
			AP Easy	71.4	# 14	Compare
			AP Medium	61.2	# 17	Compare
			AP Hard	51.2	# 15	Compare
Keypoint Detection	MS COCO	ResNet-50	Validation AP	72.2	# 10	Compare
Pose Estimation	OCHuman	ResNet-152	Test AP	33.3	# 8	Compare
Pose Estimation	OCHuman	ResNet-152	Validation AP	41.0	# 7	Compare
Keypoint Detection	OCHuman	ResNet-50	Test AP	29.5	# 7	Compare
Keypoint Detection	OCHuman	ResNet-50	Validation AP	32.1	# 7	Compare
Keypoint Detection	OCHuman	ResNet-152	Test AP	33.3	# 4	Compare
Keypoint Detection	OCHuman	ResNet-152	Validation AP	41.0	# 3	Compare
2D Human Pose Estimation	OCHuman	ResNet-50	Test AP	30.4	# 8	Compare
2D Human Pose Estimation	OCHuman	ResNet-50	Validation AP	37.8	# 6	Compare
2D Human Pose Estimation	OCHuman	ResNet-152	Test AP	33.3	# 5	Compare
2D Human Pose Estimation	OCHuman	ResNet-152	Validation AP	41.0	# 3	Compare
Pose Estimation	OCHuman	ResNet-50	Test AP	29.5	# 11	Compare
Pose Estimation	OCHuman	ResNet-50	Validation AP	32.1	# 11	Compare
Pose Tracking	PoseTrack2017	MSRA (FlowTrack)	MOTA	57.81	# 5	Compare
Pose Tracking	PoseTrack2017	MSRA (FlowTrack)	mAP	74.57	# 2	Compare
Pose Tracking	PoseTrack2018	MSRA	MOTA	61.37	# 4	Compare
Pose Tracking	PoseTrack2018	MSRA	mAP	74.03	# 1	Compare

Results from Other Papers

Task	Dataset	Model	Metric Name	Metric Value	Rank	Compare
Pose Estimation	AIC	SimpleBaseline (ResNet-152)	AP	29.9	# 8	See all
			AP50	73.8	# 4	See all
			AP75	18.3	# 5	See all
			AR	34.3	# 5	See all
			AR50	76.9	# 5	See all
Pose Estimation	AIC	SimpleBaseline (ResNet-50)	AP	28.0	# 10	See all
			AP50	71.6	# 6	See all
			AP75	15.8	# 7	See all
			AR	32.1	# 7	See all
			AR50	74.1	# 7	See all
2D Human Pose Estimation	JHMDB (2D poses only)	SimplePose	PCK	94.4	# 2	See all
Multi-Person Pose Estimation	OCHuman	SimplePose	Validation AP	24.1	# 6	See all
			AP50	37.4	# 7	See all
			AP75	26.8	# 7	See all

Methods

Add Remove

1x1 Convolution • Average Pooling • Batch Normalization • Bottleneck Residual Block • Convolution • Global Average Pooling • Kaiming Initialization • Max Pooling • ReLU • Residual Block • Residual Connection • ResNet

Edit Social Preview

Simple Baselines for Human Pose Estimation and Tracking

Code Edit Add Remove Mark official

Tasks Edit Add Remove

Datasets Edit

Results from the Paper Edit