List your company

Audio-to-Text Retrieval on ESC50

95Recall@1Proposed
91.3692.30593.2594.195Apr 25, 2026

Evaluation results

MethodLinks
ProposedModality=a2t, audio encoder=HTSAT-tiny, text encoder=RoBERTa-large2026.049599.599.897.2
Microsoft-CLAPModality=a2t2026.0493.599.810095.6
LAION-CLAPModality=a2t2026.0491.599.599.794.7