Download EVIMO

Please refer to our documentation for detailed instructions on how to read and parse the data.

The authors would appreciate if you cite our papers:

  1. EV-IMO: Motion Segmentation Dataset and Learning Pipeline for Event Cameras [bibtex] [code]
  2. Learning Visual Motion Segmentation Using Event Surfaces [bibtex]
  3. SpikeMS: Deep Spiking Neural Network for Motion Segmentation [bibtex]

DAVIS 346C

Sensor:

  • Resolution: 346 x 260
  • Datasheet: download

FOV: 70°

  • Intrinsics calibration file: download

Recordings:

  • Total recording time ~30 minutes
  • Up to 3 independently moving objects objects
  • Multiple types of scenes (varying backgrounds and motion speeds)

Evaluation Sets

Background Data Preview (Hover for animation)
  • boxes
    		•

  • Raw .bag  (0.6 GB)

  • Python .npz  (0.5 GB)

  • Text .txt  (0.7 GB)

    		•  Video
  • floor
    		•

  • Raw .bag  (0.1 GB)

  • Python .npz  (0.1 GB)

  • Text .txt  (0.1 GB)

    		•  Video
  • wall
    		•

  • Raw .bag  (1.2 GB)

  • Python .npz  (0.9 GB)

  • Text .txt  (1.4 GB)

    		•  Video
    Background Data Preview (Hover for animation)
  • table
    		•

  • Raw .bag  (0.7 GB)

  • Python .npz  (0.6 GB)

  • Text .txt  (0.8 GB)

    		•  Video
  • tabletop
    		•

  • Raw .bag  (0.3 GB)

  • Python .npz  (0.2 GB)

  • Text .txt  (0.4 GB)

    		•  Video
  • fast
    		•

  • Raw .bag  (0.6 GB)

  • Python .npz  (0.5 GB)

  • Text .txt  (0.7 GB)

    		•  Video

    Training Data

    Background Data Preview (Hover for animation)
  • boxes
    		•

  • Raw .bag  (7 GB)

  • Python .npz  (5 GB)

  • Text .txt  (9 GB)

    		•  Video
  • floor
    		•

  • Raw .bag  (3 GB)

  • Python .npz  (2 GB)

  • Text .txt  (3 GB)

    		•  Video
  • wall
    		•

  • Raw .bag  (2 GB)

  • Python .npz  (1 GB)

  • Text .txt  (2 GB)

    		•  Video
    Background Data Preview (Hover for animation)
  • table
    		•

  • Raw .bag  (4 GB)

  • Python .npz  (3 GB)

  • Text .txt  (5 GB)

    		•  Video
  • tabletop
    		•

  • Raw .bag  (3 GB)

  • Python .npz  (2 GB)

  • Text .txt  (4 GB)

    		•  Video
  • tabletop-ego
    		•

  • Raw .bag  (2 GB)

  • Python .npz  (2 GB)

  • Text .txt  (2 GB)

    		•  Video
  • sanity check
    		•

  • Raw .bag  (5 GB)

  • Python .npz  (4 GB)

  • Text .txt  (6 GB)

    		•  Video
  • calibration
    		•

  • Raw .bag  (1 GB)

    		•  Video

    Background Depth

    The files above provide the depth only for the foreground objects. To create depth for background and foreground objects please run the script on the evimo1 branch of EVIMO's GitHub with _no_bg:=false

    License

    This dataset is provided under the Creative Commons Attribution-ShareAlike 4.0 International public license (CC BY-SA 4.0). This means that if the data is redistributed, attribution must be given and the license cannot be changed. Commercial use and adaptation are allowed. A summary of the terms is available here.