List your company

Audio Captioning on AudioCaps AudioSet (test)

48.5SPIDErHTSAT-BART
40.38842.49444.646.706Mar 28, 2025

Evaluation results

MethodLinks
HTSAT-BART2025.0348.52578.718.2—
CNN14-BART2025.0346.824.775.617.9—
CLAP2023 (Reference)2025.0345.5————
M2D-CLAP2025Audio encoder AudioSet fine-tuning=true2025.034524.372.417.663.9
Cacophony2025.0344.823.672.816.8—
LAION-CLAPAudio encoder AudioSet fine-tuning=true2025.0344.623.871.717.664
M2D-CLAP2025 stage2Audio encoder AudioSet fine-tuning=false2025.0343.423.769.617.262.5
M2D-CLAP2025 stage2.1Audio encoder AudioSet fine-tuning=false2025.0343.223.669.417.162.3
M2D-CLAP2025 stage1Audio encoder AudioSet fine-tuning=false2025.0342.623.368.316.961.9
CLAP2023Audio encoder AudioSet fine-tuning=false2025.0340.722.465.216.360.5