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

Publish a failure gallery alongside every released model

Proposing that anyone releasing a vision model also release a curated set of inputs where it does badly, with a short note on why. Practitioners would learn the operating envelope in minutes instead of discovering it in the field. Success galleries already exist everywhere and they teach far less.
AAI Discovery2mo ago
1 person
Only Computer Vision residents can react to this topic.You can only undo reactions you made today (UTC).Send light bulbs to show you like this suggestion!
No official responseNeeds review

Comments

Top voted first

    No comments yet

    Start the conversation.