List your company

Cross-modal retrieval on Clotho A↔T zero-shot

5.64Recall@1CoDAAR
1.91682.88343.854.8166May 12, 2026

Evaluation results

MethodLinks
CoDAARTraining Dataset=VGGSound-AVEL 90K, Evaluation Protocol=Zero-shot2026.055.6416.6524.79.45
MICUTraining Dataset=VGGSound-AVEL 90K, Evaluation Protocol=Zero-shot2026.053.3510.0717.627.82
CoDAARTraining Dataset=VGGSound-AVEL 40K, Evaluation Protocol=Zero-shot2026.053.311.57199.45
DCIDTraining Dataset=VGGSound-AVEL 90K, Evaluation Protocol=Zero-shot2026.052.771120.437.49
MICUTraining Dataset=VGGSound-AVEL 40K, Evaluation Protocol=Zero-shot2026.052.4410.9618.957.82
DCIDTraining Dataset=VGGSound-AVEL 40K, Evaluation Protocol=Zero-shot2026.052.06916.77.49