Figure 4: Typical usage scenarios.
(a) Evaluation of a tracking algorithm. LTDB videos are provided as input to a tracking algorithm. Computed tracks can be compared with respect to the ground-truth tracks using a methodology of choice such as the complete graph comparison15,16. In the example red arrows indicate errors where a cell was detected not sufficiently close, and when a track was interrupted. (b) Machine learning dataset. LTDB videos and tracks can potentially be used in the context of supervised machine learning as training and validation dataset. The generated predictive model can be generalized and used to track new videos. (c) Resource for big data analysis. Properties of leukocyte migration in different experimental conditions can potentially be discovered by the application of pattern recognition on LTDB metadata and tracks.
