ResearchTasksVision-Language (Image Description)FollowBenchmarksDataset NameSOTA MethodSortMost resultsRecently updatedMost papersApplyDataset NameSOTA methodMetricTrendResultsLast UpdatedNine synthetic images dataset with Qwen-VL-Plus length-matched referencesMini-Omni276.09Avg CER2May 6, 2026
Nine synthetic images dataset with Qwen-VL-Plus length-matched referencesMini-Omni276.09Avg CER2May 6, 2026