Dataset Viewer
Auto-converted to Parquet
Search is not available for this dataset
video
video
YAML Metadata Warning: empty or missing yaml metadata in repo card (https://huggingface.co/docs/hub/datasets-cards)

GIST Multi-Camera Dataset

Collect and Annotate 3D Multi-Camera Multi-Object Tracking Dataset

Prepare and print a checkerboard

1) Go to https://calib.io/
2) Pattern Generator (1) Target Type: Checkerboard, 
3) Board Width [mm]: 841, (3) Board Height [mm]: 1189, 
4) Rows: 6, (5) Columns: 4, Checker Width [mm]: 195
5) Download PDF
6) Print it two piceces of the PDF with A0 paper (841x1189 mm)
7) Glue them together precisely and tightly

Install Packages

sudo apt install ffmpeg
pip install numpy==1.23.5 pandas==1.5.1 opencv-python==4.10.0.84

Process

  1. Videos are captured from Android devices using RecSync with some modifications.
    • Adding TrueTime dependency 'com.github.instacart:truetime-android:3.4' to app/build.gradle. In MainActivity.java make sure to init TrueTime, trueTimeInitialized.await();
    • CameraController.java, long synchronizedTimestampNs = TrueTime.now().getTime();. This change adapts one captured frame cooresponding to a timestamp obtained from TrueTime.
    • SoftwareSyncController.java, leaderAddress = InetAddress.getByName("172.17.3.136"); change IP address of leader server.
    • If all Android devices are not the same model, we need to move some statements out of if (isLeader) {})
      exposureSeekBar.setVisibility(View.VISIBLE);
      sensitivitySeekBar.setVisibility(View.VISIBLE);
      exposureSeekBar.setOnSeekBarChangeListener(
      
      This setting allows us to ajust sensitivity and exposure of Android camera.
  2. Videos are extracted using FFMPEG
  3. Frames of each video are matched using timestamp obtained from TrueTime
  4. calibration_syncvideos.py outputs calibration matrix from synchronized and extracted frames. Synchronized frames are saved in a folder name as ./GMC0715/S20.

Contact

Linh Ma (linh.mavan@gm.gist.ac.kr), Machine Learning & Vision Laboratory, GIST, South Korea

Citation

If you find this project useful in your research, please consider citing by:

@article{linh2024inffus,
      title={Track Initialization and Re-Identification for {3D} Multi-View Multi-Object Tracking}, 
      author={Linh Van Ma, Tran Thien Dat Nguyen, Ba-Ngu Vo, Hyunsung Jang, Moongu Jeon},
      journal={Information Fusion},
      volume = {111},
      year={2024},
      publisher={Elsevier}
}

license: mit

Downloads last month
53