ResearchBenchmarksCross-modal Retrieval (Image to Audio) on iNaturalist Species-level Retrieval seenCross-modal Retrieval (Image to Audio) on iNaturalist Species-level Retrieval seenFollow57.5Top-1 AccuracyBioVITA-2.289613.232728.75544.2773Mar 25, 2026Evaluation resultsMethodMethodLinksTop-1 AccuracyTop-5 AccuracyBioVITATraining Stage=Stage2Training Stage=Stage22026.0357.585.6BioVITATraining Stage=Stage1Training Stage=Stage12026.0348.678.8TaxaBind2026.0316.241.4ImageBind2026.032.912.2Random2026.030.010.05