ResearchBenchmarksText-to-sound generation on AudioCaps (val)Text-to-sound generation on AudioCaps (val)Follow4.307RelevanceGT2.68463.10583.5273.9482Jul 20, 2022Evaluation resultsMethodMethodLinksRelevanceFidelityIntelligibilityMOSGTDescription=Ground Tru...Description=Ground Truth sound2022.074.3074.1673.8734.116DiffsoundModel Type=Diffusion-b...Model Type=Diffusion-based2022.073.8333.4873.363.56ARModel Type=AutoregressiveModel Type=Autoregressive2022.072.7472.72.9132.786