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

RoboVQA

Benchmarks

Task NameDataset NameSOTA ResultTrend
Long-horizon reasoning for robotic manipulationRoboVQA
B-1 Score75.1
28
Temporal Task PlanningRoboVQA
Score74.5
20
Next-Step-Prediction Style PlanningRoboVQA
Performance Score64.52
16
Robotic Video Question AnsweringRoboVQA
Score64.6
14
Embodied Question AnsweringRoboVQA
BLEU-177.1
13
Robotic Video Question AnsweringRoboVQA (test)
BLEU-172.7
6
Video Question AnsweringRoboVQA
BLEU-186.97
5
Embodied QARoboVQA (test)
BLEU-177.1
5
Robotic ReasoningRoboVQA (val)
BLEU-442.8
4
Showing 9 of 9 rows