Visit the work webpage.

We propose an 80-hour Dataset of Multimodal Semantic Egocentric Videos (DoMSEV) covering a wide range of activities. The videos were recorded using either a GoPro Hero camera or a built setup composed of a 3D Inertial Movement Unit (IMU) attached to the Intel Realsense R200 RGB-D camera. Different people recorded the videos in varied illumination and weather conditions. The recorders labeled the videos informing the scene where some segment were taken (e.g., indoor, urban, crowded environment, or nature), the activity performed (walking, running, standing, browsing, driving, biking, eating, cooking, eating, observing, in conversation, playing, or shopping), if something caught their attention and when they interacted with some object. Also, we create a profile for each recorder representing their preferences over a set of objects and visual concepts.


  title     = {A Weighted Sparse Sampling and Smoothing Frame Transition Approach for Semantic Fast-Forward First-Person Videos},
  booktitle = {2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},  
  author    = {Silva, Michel and Ramos, Washington and Ferreira, João and Chamone, Felipe and Campos, Mario and Nascimento, Erickson R.},
  Year      = {2018},
  Address   = {Salt Lake City, USA},
  month     = {Jun.},
  pages     = {2383-2392},
  doi       = {10.1109/CVPR.2018.00253},
  ISBN      = {978-1-5386-6420-9}

Videos info

Detailed videos information such as Duration, Resolution, Capture Device, FOV, FPS, Camera Mounting, and Sensors (GPS, IMU, Depth).

Copyright Notice

This dataset is published under the Creative Commons Attribution-NonCommercial 4.0 International (CC BY-NC 4.0) License. This means you must give appropriate credit, provide a link to the license, and indicate if changes were made. You may do so in any reasonable manner, but not in any way that suggests the licensor endorses you or your use. You may not use the material for commercial purposes.


  • 3D built Model
  • 3D model of a case for the Realsense R200 RGB-D camera with support for the LORD MicroStrain 3DM-GX3-25 and GoPro mount adapter.

    Download the STL model of the setup.

  • DoMSEV – Dataset of Multimodal Semantic Egocentric Videos

    You can customize the download following the next steps:

    1. Select files to download (expand the tree for more detail).
    2. After selecting at least one file the Download buttons will be enabled.

    Then, you can either click in the

    • Download sh file buttom to download a shell script file with the wget command to download all the selected files and place them in the repective folders.
    • Download files list buttom to download a list of files that you can use into a download manager software.

    Total size: 0 Kb
    Back to the project page.