ResearchTasksContextual Image GenerationFollowBenchmarksDataset NameSOTA MethodSortMost resultsRecently updatedMost papersApplyDataset NameSOTA methodMetricTrendResultsLast UpdatedUNIVL-ImgGen single and multiple masksUNIVL + T510.49FID10May 22, 2026VisDial 1.0 (test)GILL0.645CLIP Similarity9Feb 26, 2026VIST 28 (test)GILL0.641CLIP Similarity7Feb 26, 2026COCO 2017 (val)UniVL16.99FID3May 22, 2026