48 datasets found

Human action clips and segments dataset for recognition and temporal localization

The Penn Action dataset is a real video dataset of people performing various indoor and outdoor sports with annotations of human joint locations.

The MSRActionPairs Dataset consists of 12 action types performed by 10 subjects. Each subject performs every action three times. This dataset contains 6 pairs of similar actions...

The MSR Action3D dataset covers 20 sports action types and 10 subjects. All subjects perform each action two or three times. The dataset is very challenging due to the high...

The UTKinect-Action dataset consists of 10 action types performed by 10 subjects. All subjects perform each action twice. Since subjects are free to move in the environment, the...

UTD-MHAD dataset for human action recognition

The AID dataset is a benchmark for scene classification in remote sensing. It contains aerial images with 30 scene types.

The NTU-RGBD dataset is a large-scale dataset for 3D human activity analysis, containing 56,000 videos and 60 actions performed by 40 people from 80 different views.

The SBU kinect dataset is an interaction dataset acquired using the Microsoft kinect sensor.

A real-world skeleton-based human action recognition dataset.

A real-world skeleton-based human action recognition dataset.

THUMOS Challenge: Action Recognition with a Large Number of Classes.

The Charades dataset is a dataset for human action recognition. It contains 200 videos with 3,800+ action instances.

The proposed model is designed to learn recognizing actions performed by a subject in the form of mesh video.

The dataset used in this paper for Human Action Recognition tasks (HAR) using 3D Convolutional Neural Networks.

The dataset used in this paper for Human Action Recognition tasks (HAR) using 3D Convolutional Neural Networks.

The UCF-101 and HMDB-51 datasets are used for human action recognition.

The NTU-120 RGB+D dataset is the largest indoor dataset for 3D human activity analysis. It contains 114,480 action samples over 120 action classes, performed by 106 distinct...

UCF-101 is a large-scale dataset of human actions in videos.

The HACS dataset contains 200 action classes with 504K videos.