Curation workflows

Open Model ▸ Curation workflows… in the top bar. A workflow is a list of custom functions you can chain, so that when you load predictions from a model’s output a few of the steps that follow run for you. The logic is always the same: Filter trials first, then every later step runs over all the filtered trials.

If you would like a workflow step added, please raise an issue.

The steps

Step	What it does
Filter trials	Sets the trials table’s column filters, stored by column name so the same workflow runs on any session with that metadata. Every later step runs over exactly the trials the table then shows (see Trial metadata).
Predict onsets	Runs a trained LightGBM model over those trials, filling classes they do not already carry (see LightGBM for point events). The classes it wrote are passed to a following Set curation scope step whose class list is empty. Predictions from the eto.segment and eto.spot modules are not run here: load them via File ▸ Import predictions.
Set curation scope	Drops label classes into the Curation section’s scope area and picks the curation mode.
Label grid view	Opens the frame grid on the scope, from the chosen cameras, laid out and generated as configured. The workflow waits until the grid is closed.
Video grid	Opens the clip player on the scope, from the chosen cameras. The workflow waits until the grid is closed.
Segment review	Starts the label-by-label review over the scope: each label plays, two clicks re-place it, N curates. The workflow waits until the review finishes or is stopped; with nothing left to review it is skipped.
Frame-by-frame review	Starts the boundary review over the scope. The workflow waits until the review finishes or is stopped; with nothing left to review it is skipped.
Curate trials’ labels	Marks every automated label of the chosen classes, in the chosen trials (current, all, filtered or hidden), as curated. An empty class list means whatever the curation scope holds. Curating is not undoable.
Delete trials’ labels	Deletes every label of the chosen classes in the chosen trials, whatever its labeling method.
Purge short labels	Deletes state-interval labels of the chosen classes shorter than a threshold, in the chosen trials. Point events are never touched.
Stitch labels	Merges same-class state labels of one individual separated by less than a gap, in the chosen trials — the changepoint correction’s own stitch, on its own. Point events are never touched.
Correct changepoints	Presses the Changepoints tab’s manual correction — purge, stitch, snap to changepoints, purge again — with the settings that tab shows, over the current trial or every filtered trial (see Changepoint correction). Those settings are the tab’s, not the step’s, so set them there first.
Score trials against the model	Presses Score now in Model ▸ Curator feedback: each trial’s curated labels against what its run predicted, an F1 per trial into the metadata table. Flagging from those scores is the reviewer’s decision in the histogram, never a step (see Curator feedback).
Save labels	Writes the labels TSV, exactly as Ctrl+S does. Nothing reaches disk before this step.