Dataset of Manufacturing Tasks - DMT22
<p>Description: The dataset of manufacturing tasks (DMT22), features five representative assembly primitive actions (PickUp, Place, Screw, Idle and Hold), collected from 5 different subjects featuring 3 different objects, a Wii remote, a hard disk drive, and an electric screwdriver.</p> <p>Available Data: The DMT22 dataset contains recorded data from an event camera (DAVIS 240C), a depth camera (Intel RealSense D435) and a magnetic tracking system (Polhemus Liberty).</p> <ul> <li>The event camera captured events and APS frames with a resolution of 240 pixels × 180 pixels. DAVIS APS data are captured at 30 fps (frames per second). Event data is available in the original recorded format, .aedat, together with a .txt file containing label and its corresponding timestamp range, and a .mat file which contains the information from the events (coordinates x and y, timestamp and polarity).</li> <li>The electromagnetic tracker was used to capture pose data, at 20 Hz, from a sensor attached to the subject’s<br> wrist. The data is available in .mat format. It contains 6DOF data (position and orientation) followed by the timestamp of measured data (relative to the start of the recording).</li> <li>RGB-D camera data is available, at 30fps, in the original recorded format, .bag.</li> </ul> <p>The dataset features a total of 72 recordings: 4 right-handed subjects × 3 tasks × 4 performances + 1 left-handed subject × 2 configurations (left and right) × 3 tasks × 4 performances. Each task recording lasts approximately 25 seconds.</p> <p> </p>
ShareScore
40/100
Overall dataset sharing score
Score breakdown
These five areas show where the dataset supports — or may limit — practical reuse.
- Stewardship
- 8
- Harmonization
- 4
- Access
- 16
- Reuse readiness
- 8
- Engagement
- 4