Pose Estimation

1351 papers with code • 28 benchmarks • 114 datasets

Pose Estimation is a computer vision task where the goal is to detect the position and orientation of a person or an object. Usually, this is done by predicting the location of specific keypoints like hands, head, elbows, etc. in case of Human Pose Estimation.

A common benchmark for this task is MPII Human Pose

( Image credit: Real-time 2D Multi-Person Pose Estimation on CPU: Lightweight OpenPose )

Benchmarks

Add a Result

These leaderboards are used to track progress in Pose Estimation

Dataset	Best Model	Compare
MPII Human Pose	PCT (swin-l, test set)	See all
COCO test-dev	ViTPose (ViTAE-G, ensemble)	See all
Leeds Sports Poses	OmniPose	See all
OCHuman	ViTPose (ViTAE-G, GT bounding boxes)	See all
CrowdPose	BUCTD-W48 (w/cond. input from PETR, and generative sampling)	See all
MS COCO	OmniPose (WASPv2)	See all
AIC	Hulk(Finetune, ViT-L)	See all
ITOP front-view	AdaPose	See all
ITOP top-view	DECA-D3	See all
UPenn Action	OmniPose	See all
J-HMDB	SimpleBaseline + HANet	See all
MPII Single Person	4xRSN-50	See all
COCO val2017	MogaNet-B (384x288)	See all
300W (Full)	SPIGA	See all
DensePose-COCO	Parsing R-CNN + ResNext101	See all
FLIC Elbows	Stacked Hourglass Networks	See all
FLIC Wrists	Stacked Hourglass Networks	See all
UAV-Human	AlphaPose	See all
BRACE	HRNet fine-tuned on BRACE	See all
COCO minival	MSPN	See all
3DPW	HybridCap	See all
MPII	OmniPose (WASPv2)	See all
ApolloCar3D	GSNet	See all
Pix3D	Mid-Level based	See all
KITTI 2015	GeoNet	See all
MERL-RAV	SPIGA	See all
MS-COCO	UniHCP (finetune)	See all
COCO 2017 val	MogaNet-S (384x288)	See all

Show all 28 benchmarks

Collapse benchmarks

Libraries

Use these libraries to find Pose Estimation models and implementations

open-mmlab/mmpose

32 papers

5,022

PaddlePaddle/PaddleDetection

9 papers

12,085

DeepLabCut/DeepLabCut

6 papers

4,293

osmr/imgclsmob

6 papers

2,916

Datasets

Subtasks

6D Pose Estimation

Hand Pose Estimation

6D Pose Estimation using RGB

Multi-Person Pose Estimation

Head Pose Estimation

Human Pose Forecasting

Animal Pose Estimation

6D Pose Estimation using RGBD

Vehicle Pose Estimation

RF-based Pose Estimation

Car Pose Estimation

Hand Joint Reconstruction

Activeness Detection

Semi-supervised 2D and 3D landmark labeling

Most implemented papers

Most implemented Social Latest No code

3D Bounding Box Estimation Using Deep Learning and Geometry

smallcorgi/3D-Deepbox • • CVPR 2017

In contrast to current techniques that only regress the 3D orientation of an object, our method first regresses relatively stable 3D object properties using a deep convolutional neural network and then combines these estimates with geometric constraints provided by a 2D object bounding box to produce a complete 3D bounding box.

Paper
Code

Lifting from the Deep: Convolutional 3D Pose Estimation from a Single Image

DenisTome/Lifting-from-the-Deep-release • • CVPR 2017

We propose a unified formulation for the problem of 3D human pose estimation from a single raw RGB image that reasons jointly about 2D joint estimation and 3D pose reconstruction to improve both tasks.

Paper
Code

PoseCNN: A Convolutional Neural Network for 6D Object Pose Estimation in Cluttered Scenes

yuxng/PoseCNN • • 1 Nov 2017

We conduct extensive experiments on our YCB-Video dataset and the OccludedLINEMOD dataset to show that PoseCNN is highly robust to occlusions, can handle symmetric objects, and provide accurate pose estimation using only color images as input.

Paper
Code

Real-time 2D Multi-Person Pose Estimation on CPU: Lightweight OpenPose

osmr/imgclsmob • • 29 Nov 2018

In this work we adapt multi-person pose estimation architecture to use it on edge devices.

Paper
Code

3D human pose estimation in video with temporal convolutions and semi-supervised training

facebookresearch/VideoPose3D • • CVPR 2019

We start with predicted 2D keypoints for unlabeled video, then estimate 3D poses and finally back-project to the input 2D keypoints.

Paper
Code

Learning from Simulated and Unsupervised Images through Adversarial Training

carpedm20/simulated-unsupervised-tensorflow • • CVPR 2017

With recent progress in graphics, it has become more tractable to train models on synthetic images, potentially avoiding the need for expensive annotations.

Paper
Code