# Curation: Post-Annotation Labels are rarely perfect on the first pass. After annotation, use FiftyOne to verify label quality and prepare your dataset for training: find likely [annotation mistakes](../brain/index.md#brain-label-mistakes), review labels as [object patches](../user_guide/using_views.md#object-patches-views), [filter and slice](../user_guide/app.md#app-filtering) your dataset in the App, and audit samples with the FiftyOne Enterprise Data Quality 🚀 panel.