3D Hand Reconstruction(一) 数据

3D Mesh 数据：

1、FreiHAND 《Freihand: A dataset for markerless capture of hand pose and shape from single
rgb images》 **MANO

RGB image (224x224 pixels)

Hand segmentation mask (224x224 pixels)

Intrinsic camera matrix K

Hand scale (metric length of a reference bone)

3D keypoint annotation for 21 Hand Keypoints

3D shape annotation

Computer Vision Group, Freiburg
GitHub - lmb-freiburg/freihand: A dataset for estimation of hand pose and shape from single color images.

2、HO3D《Honnotate: A method for 3d annotation of hand and object poses》**MANO

交互

ICG - Hand-Object 3D Pose Annotation

GitHub - anilesec/ho3d: Creating Meshes and SDFs

3、HIC 《Capturing hands in action using discriminative salient points and physics simulation》

RGB-D image, 3D object shape, and MANO-fitted handshape.

Capturing Hands in Action using Discriminative Salient Points and Physics Simulation

4、ObMan 《Learning joint reconstruction of hands and manipulated objects.》

RGB-D image, 3D hand joints, 3D hand mesh, object mesh, camera pose parameters

ObMan Dataset

https://drive.google.com/drive/folders/1L-9QmiD3SqjZ2jf4DvtwD4LMQRkzl0cM

5、YT-3D 《Weakly supervised mesh-convolutional hand reconstruction in the wild》

3D joint and mesh annotation.

https://github.com/arielai/youtube_3d_hands

 6、Monocular Total Capture (MTC)

《Monocular total capture: Posing face, body, and hands in the wild》

Monocular Total Capture - CMU Panoptic Dataset

全身mesh

虚拟数据：

《3D Hand Shape and Pose Estimation from a Single RGB Image》

https://github.com/3d-hand-shape/hand-graph-cnn/tree/master/data

3D joint location数据：

1、《Learning to Estimate 3D Hand Pose from Single RGB Images》

Rendered Handpose Dataset（RHD） 

It contains 41258 training and 2728 testing samples. Each sample provides:

	-	RGB image (320x320 pixels)
	-	Depth map (320x320 pixels)
	-	Segmentation masks (320x320 pixels) for the classes: background, person, three classes for each finger and one for each palm
	-	21 Keypoints for each hand with their uv coordinates in the image frame, xyz coordinates in the world frame and a visibility indicator
	-	Intrinsic Camera Matrix K

https://lmb.informatik.uni-freiburg.de/data/RenderedHandpose/RHD_v1-1.zipComputer Vision Group, Freiburghttps://lmb.informatik.uni-freiburg.de/data/RenderedHandpose/RHD_v1-1.zip

2、《A hand pose tracking benchmark from stereo matching》

Stereo Hand Pose tracking benchmark（STB）

GitHub - zhjwustc/icip17_stereo_hand_pose_dataset: A hand pose tracking benchmark from stereo matching[dataset]

3、STEREO 

《3d hand pose tracking and estimation using stereo matching》

4、EXTER+OBJECT

《Real-time joint tracking of a hand manipulating an object from rgb-d input》

5、EGODEXTER

《Real-time hand tracking under occlusion from an egocentric rgb-d sensor》