tao-route-visual-changenet-samples

To route Visual ChangeNet samples after gap analysis, you process the output samples based on label eligibility, directing them into subsets suitable for k-NN Mining or AnomalyGen augmentation modules within a VCN AOI SDA pipeline.

VCN sample subsetting requires a source_pool_csv for k-NN Mining and a defined set of supported labels to determine AnomalyGen eligibility, ensuring samples are accurately routed for augmentation.

Label eligibility determines routing by checking each VCN sample against supported labels, directing qualifying samples to AnomalyGen while allocating others using the source_pool_csv for k-NN Mining.

Yes, you can use pandas to manage VCN sample routing, as the processing logic relies on Python and utilizes pandas to handle the data structures required for routing samples into appropriate subsets.

You should route VCN samples into subsets immediately after performing gap analysis using tao-analyze-gaps-visual-changenet, ensuring the data is properly prepared before entering the augmentation phase.