User guide
Data is imported, and features capturing biological behaviours are engineered; annotation is then performed either from simulated ground truth, direct human annotation, or perturbation of entire datasets; a subset of features expected to capture the behaviours are then selected and used to train an ML model; the trained model is used to annotate unseen datasets, and the resulting segmentations are used to analyse biological states, and identify unexpected relationships.
The DeepTRACE workflow
The pages below take you chronologically through using the software.
- Loading new data
- Feature engineering
- Human annotation of tracks
- Importing simulated ground truth
- Training a model
- Classifying new data
- Analysing results
- Discovery of relationships
Continue to: Loading new data →