List your company

Audio Captioning on AudioCaps

80.3CIDErEnCLAP-large
25.429639.674853.9268.1652Apr 17, 2023Oct 22, 2023Apr 28, 2024Nov 3, 2024May 11, 2025Nov 16, 2025May 24, 2026

Evaluation results

MethodLinks
EnCLAP-large2024.0180.325.518.849.5————
CoDi2023.0578.9—18.248————
CoDi2023.1278.9———————
HTSAT-BART2025.0378.72518.248.5————
HTSAT-BART2023.0578.725———28.350.7—
Method [18]2023.0578.7———————
WavCaps2024.0178.72518.248.5————
SOTA2024.0678.7———————
VAST2023.0578.2———————
VAST2023.0578.124.7———29.550.9—
EnCLAP-base2024.017824.718.648.3————
AL-MixGen2024.0176.924.218.147.5————
CNeXt-Trans2024.0176—18.247.1————
CNN14-BART2025.0375.624.717.946.8————
CNN14-BART2023.0575.624.7———27.249.9—
AL-MixGen2023.0575.5—17.7—————
BART-Finetune2023.0575.3—16.646————
BART-tags2024.0175.324.117.646.5————
Pengi2024.0175.223.218.246.7————
VALOR2023.0574.1———————
VALOR-B2023.0574.123.1———2749.4—
VALOR_B2023.0474.123.1———2749.4—
AF-Next-Instruct2026.0474———————
CNN14-GPT22024.0173.32417.745.5————
Cacophony2025.0372.823.616.844.8————
M2D-CLAP2025Audio encoder AudioSet fine-tuning=true2025.0372.424.317.64563.9———
LAION-CLAPAudio encoder AudioSet fine-tuning=true2025.0371.723.817.644.664———
Audio Flamingo 32026.0470———————
M2D-CLAP stage2Audio encoder AudioSet fine-tuning=false2025.0369.623.717.243.462.5———
M2D-CLAP stage2.1Audio encoder AudioSet fine-tuning=false2025.0369.423.617.143.262.3———
M2D-CLAP stage1Audio encoder AudioSet fine-tuning=false2025.0368.323.316.942.661.9———
LHDFF2024.016823.217.142.6————
ACT2024.0167.922.21642————
ACT2023.0467.922.2———25.246.8—
Liu et al.2023.0466.723.2———25.148—
MiCo2024.0666.2———————
Xu et al.2023.056622.9———23.146.7—
Xu et al. [115]2023.046622.9———23.146.7—
CLAP2023Audio encoder AudioSet fine-tuning=true2025.0365.222.416.340.760.5———
AudioCaps2023.0559.3—14.436.9————
Kim et al.2023.0459.320.3———21.945—
Audio Flamingo# Shots=162024.0254.6———————
Audio Flamingo# Shots=82024.0253.8———————
Audio Flamingo# Shots=42024.0251.8———————
LTBM (local)Keep Ratio=0.752026.0549.51———————
Global Top-KKeep Ratio=0.752026.0549.12———————
UIO-2XXLModel Size=XXL2023.1248.9———————
BaselineKeep Ratio=0.752026.0548.25———————
BaselineKeep Ratio=0.52026.0548.25———————
BaselineKeep Ratio=0.252026.0548.25———————
Global MergeKeep Ratio=0.752026.0548.23———————
Global MergeKeep Ratio=0.52026.0548.05———————
LTBM (local)Keep Ratio=0.52026.0547.82———————
Segmentwise Top-KKeep Ratio=0.752026.0547.74———————
UniAvgKeep Ratio=0.52026.0547.48———————
Segmentwise Top-KKeep Ratio=0.52026.0547.28———————
Global Top-KKeep Ratio=0.52026.0545.78———————
UIO-2LModel Size=L2023.1245.7———————
UIO-2XLModel Size=XL2023.1245.7———————
LTBM (local)Keep Ratio=0.252026.0539.09———————
Global MergeKeep Ratio=0.252026.0538.09———————
RECAP# Shots=42024.0235.9———————
Segmentwise Top-KKeep Ratio=0.252026.0532.67———————
Global Top-KKeep Ratio=0.252026.0529.57———————
ImageBindZero-shot=true2023.1227.8———————
UniAvgKeep Ratio=0.252026.0527.54———————
Captioning-scratchLabel=Caption, Evaluation Protocol=audio-language alignment2026.03———————43.58
Captioning-scratch (Ours)Label=Caption*, Evaluation Protocol=audio-language alignment2026.03———————45.07
CLAP2023 (Ref)2025.03———45.5————
Contrastive-scratchLabel=Caption, Evaluation Protocol=audio-language alignment2026.03———————44.5
Contrastive-scratch (Ours)Label=Caption*, Evaluation Protocol=audio-language alignment2026.03———————44.54
Kimi-Audio-InstructDecoder Size=7B2025.08————49———
MiDashengLMDecoder Size=7B2025.08————62.18———
MTC (AudioSet)Label=Tag, Evaluation Protocol=audio-language alignment2026.03———————46.67
MTC (Ours-UTS)Label=Tag*, Evaluation Protocol=audio-language alignment2026.03———————44.4
Multi-Task (Ours)Label=Tag*, Caption*, Evaluation Protocol=audio-language alignment2026.03———————45.17
PAR (Ours-UTS)Label=Tag Sequence*, Evaluation Protocol=audio-language alignment2026.03———————44.8
Pengi2023.05———46.67————
Qwen2.5-OmniDecoder Size=7B2025.08————60.79———