Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

RAVEN-QA

Benchmarks

Task NameDataset NameSOTA ResultTrend
Video Question AnsweringRAVEN-QA Hard Queries
Accuracy96.8
64
Video Question AnsweringRAVEN-QA Simple Queries
Accuracy100
64
Navigation Question AnsweringRAVEN-QA Habitat Simulation Tasks
Overall Accuracy88.75
28
Navigation Question AnsweringRAVEN-QA Real-World Robot Tasks
Overall Accuracy97.14
28
Showing 4 of 4 rows