Skip to content

Datasets

Hugging Face datasets with plain-language descriptions: what's inside and what models to train on them. 4,788 in the catalog so far, with new ones added every hour.

Data type

For which models

Find a dataset for your goal

Answered by Gemini 4 · the chat is kept only in this tab

Datasetropedia-ai/xperience-10mTop pick

Egocentric multimodal experience data

A large corpus of human action video, audio, and 3D data for robotics and embodied AI.

  • Video classification
  • Image → text
  • Depth estimation
  • English
43K 2571M–10M rows🔒 on request
DatasetHuggingFaceFV/finevideo

Videos with descriptions and questions

An English video corpus for training models to answer questions about videos.

  • Visual question answering
  • Video text → text
  • English
42K 41110K–100K rows🔒 on request
Datasetgagandeepreehal/indus-ad-monocam

Monocular car camera recordings on Indian roads

96 single-camera driving videos and over 105 thousand frames with depth and semantic segmentation annotations.

  • Robotics
  • Depth estimation
  • Image segmentation
  • English
40.6K 1100K–1M rows🔒 on request
Datasetgroundedsi/RoboCap-10kTop pick

RoboCap-10k: robotics dataset with 3D poses

10,000 hours of real human work video with head and hand pose annotations in video format for robotics.

  • Video classification
  • Depth estimation
  • Text → video
  • English
39.5K 81M–10M rows🔒 on request
DatasetVeroLM/VeroMotion

Short Video Clips with Active Object Motion

A filtered selection of 10,000 to 100,000 dynamic video clips without static scenes for training world models and video generation.

    38.9K 110K–100K rows🔒 on request
    DatasetMutonix/Vript

    Video Clips with Script Annotations and Camera Motion

    About 400k English video clips with detailed scene, shot, and camera motion descriptions.

    • Video classification
    • Visual question answering
    • Text → video
    • English
    37.3K 27100K–1M rows
    DatasetX-Humanoid/RoboGene

    Synthetic video data for RoboGene robots

    A video dataset generated by an agent framework for pre-training Vision-Language-Action models.

      35.4K 0up to 1K rows
      DatasetOpenGalaxea/Galaxea-Open-World-DatasetTop pick

      Household mobile manipulation robot videos

      Over 500 hours of R1-Lite mobile robot video with action annotations for robotics training.

      • English
      • Chinese
      35.1K 58over 1T rows🔒 on request
      Dataset2077AIDataFoundation/TouchScaleTop pick

      Multimodal visual-tactile interaction dataset

      100 hours of egocentric camera and tactile glove recordings for robotic manipulation training.

      • Robotics
      • Video classification
      • English
      34.5K 2110K–100K rows🔒 on request
      Datasethffordata/h3-fewstep-benchmark-20260920

      MiniMax-H3 audio-visual video generation benchmark

      4,320 audio-visual videos from 15 accelerated versions of the MiniMax-H3 model for evaluating generation quality in few steps.

      • Text → video
      30.5K 1
      DatasetCharly-chan/hoigen-filtered-videos

      Filtered videos from the HOIGen-1M dataset

      A selection of 28,562 video clips filtered via an allowlist from the large HOIGen-1M dataset.

      • Other
      • English
      29.7K 01K–10K rows
      Datasetshpouladi/American-Sign-Language-Dataset

      American Sign Language (ASL) video dictionary

      Over 108,000 short video clips across 2,208 American Sign Language words for sign recognition.

        28.9K 0up to 1K rows