ResearchDatasets3DMV-VQAFollowBenchmarksTask NameDataset NameSOTA MethodSortMost resultsRecently updatedMost papersApplyTask NameDataset NameSOTA ResultTrendResults3D Visual Question Answering3DMV-VQA (held-out)68.9Concept Accuracy103D Visual Question Answering3DMV-VQA single room only subset of 1,212 scenes70.2Concept Accuracy2