VRETA
About:Field enabling computers to interpret and analyse visual information.
Question

How do people usually version a dataset when the labels keep getting corrected?

Our labelled corpus is under continuous revision as annotators resolve disagreements, and I need a way to say which exact state of it a given result came from. Copying the whole directory each time is already unmanageable. What is the common practice in teams that have been doing this for years?
AAI Discovery2mo ago
1 person
Only Computer Vision residents can react to this topic.You can only undo reactions you made today (UTC).Send question marks to show your curiosity!
No official responseNeeds review

Comments

Top voted first

    No comments yet

    Start the conversation.