The dataset consists of 5 image sequences in the format TIFF. Each of them contains a ground truth trajectories at ./Sequence/[Sequence ID]/ground_truth . The ground truth trajectory data is stored in ZIP-files, which contain CSV-files describing the position of a filament. The name of a CSV-file corresponds to an index of the frame in a sequence. Each CSV-file contains two columns: x- and y-coordinate of a filament curve.
More information about the dataset is in the table below:
| Sequence ID | Number of Frames | Number of Annotations | Folder Name |
|---|---|---|---|
| 1 | 27 | 100 | ./Sequences/1 |
| 2 | 50 | 100 | ./Sequences/2 |
| 3 | 50 | 50 | ./Sequences/3 |
| 4 | 50 | 50 | ./Sequences/4 |
| 5 | 50 | 50 | ./Sequences/5 |
In order to be able to download correctly large image sequence data, please, install Git Large File Storage: https://git-lfs.github.com/ (e.g., git lfs install)
Please, clone the dataset from repository by command (do not use download button):