8 datasets found

PandaNet is a single-shot approach for multi-person 3D pose estimation that can handle images with a large number of people, strong occlusions, and varying sizes.

The CVUSA and CVACT datasets are used for cross-view geolocalization. The VIGOR dataset is used for cross-view image retrieval and 3-DoF pose estimation.

The CVUSA and CVACT datasets are used for cross-view geolocalization. The VIGOR dataset is used for cross-view image retrieval and 3-DoF pose estimation.

MuPoTS-3D is a representative dataset for monocular 3DMPPE, composed of 20 few-second-long sequences with 2–3 people interacting with each other.

The HO-3D dataset consists of 78,000 images frames on 10 objects.

A new dataset for multi-person 3D pose estimation with occlusion handling in moving camera environments.

3D human pose and mesh estimation, the task of recon-structing human pose in 3D space given a 2D image of the person, is an ill-posed problem, and many data-driven ap-proaches...

Estimating 3D human pose and shape from monocular images is a crucial task for various applications such as performance retargeting, virtual avatars, and human action recognition.