Detecting Multimodal Situations with Insufficient Context and Abstaining from Baseless Predictions

Open in new window