25 datasets found

Tags: human action recognition

Filter Results

  • Penn Action

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

  • UTD-MHAD dataset

    UTD-MHAD dataset for human action recognition

  • AID dataset

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

  • SBU kinect dataset

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

  • Charades dataset

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

  • X3D-M

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

  • Slowonly

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

  • HACS

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

  • Epic-Kitchens-100

    The Epic-Kitchens-100 dataset contains 97 verb and 300 noun classes with actions defined by the combination of nouns and verbs.

  • Something-Something-V1 and V2

    The Something-Something-V1 and V2 dataset contains 174 human action categories with 108K and 220K videos.

  • UCF101: A Dataset of 101 Human Actions Classes from Videos in the Wild

    The authors used the UCF101, HMDB51, and Diving48 datasets to evaluate the performance of their proposed algorithm.

  • Pose-Aware Video Transformers

    Human perception of surroundings is often guided by the various poses present within the environment. Many computer vision tasks, such as human action recognition and robot...

  • K400

    The dataset used in this paper is K400, a dataset for human action recognition.

  • SSv2

    The dataset used in this paper is SSv2, a dataset for human action recognition.

  • DHA

    The DHA database is orgnized with 483 depth video sequences for 23 actions.

  • UTD-MHAD

    The UTD-MHAD is consists of 861 samples of 8 subjects. There are 27 actions in total, and every subject performed each action 4 times.

  • HMDB-51

    Motion has shown to be useful for video understanding, where motion is typically represented by optical flow. However, computing flow from video frames is very time-consuming....

  • Kinetics-400

    Motion has shown to be useful for video understanding, where motion is typically represented by optical flow. However, computing flow from video frames is very time-consuming....

  • JHMDB51

    A video dataset of 51 human actions classes.

  • AVA-Kinetics

    The AVA-Kinetics dataset is a video dataset of localized human actions.