List your company

Text-to-Audio Retrieval on Clotho

0.212R@1WavCaps-CNN14
-0.008480.048760.1060.16324Oct 20, 2023Mar 26, 2024Aug 31, 2024Feb 5, 2025Jul 13, 2025Dec 18, 2025May 26, 2026

Evaluation results

MethodLinks
WavCaps-CNN14Year=2023, Setting=ZS2026.050.212———
MGA-CLAPYear=2024, Setting=ZS2026.050.208———
FLAP fusionYear=2024, Setting=ZS2026.050.203———
CacophonyYear=2024, Setting=ZS2026.050.202———
M2D-CLAPYear=2025, Setting=ZS2026.050.201———
WavCaps-HTSATYear=2023, Setting=ZS2026.050.195———
GLAPYear=2025, Setting=ZS2026.050.194———
OmniRetriever-7BYear=2026, Setting=ZS2026.050.191———
LAION-CLAP fusionYear=2023, Setting=ZS2026.050.182———
PMRLSetting=Zero-shot2025.070.168—0.44—
LanguageBindSetting=Zero-shot2025.070.167—0.52—
MS-CLAPYear=2023, Setting=ZS2026.050.167———
LanguageBind (CLIP-H/14)Year=2024, Setting=ZS2026.050.167———
GRAMSetting=Zero-shot2025.070.159—0.436—
ProposedModality=t2a, audio encoder=HTSAT-tiny, text encoder=RoBERTa-large2026.040.1580.420.5430.267
Microsoft-CLAPModality=t2a2026.040.1560.3850.510.255
LAION-CLAPModality=t2a2026.040.1460.3490.4470.231
VASTSetting=Zero-shot2025.070.124—0.364—
CyCLAPSsentence embedding training=unsupervised (Ls)2023.100.12080.34080.4676—
CLAPSsentence embedding training=unsupervised (Ls)2023.100.11920.32540.4547—
CyCLAP2023.100.1190.34950.4861—
CLAP2023.100.11670.3380.471—
MMTYear=2022, Setting=FT2026.050.067———
Omni-Embed-NemotronYear=2025, Setting=ZS2026.050.064———
ImageBindSetting=Zero-shot2025.070.06—0.284—
ImageBind (ViT-H)Year=2023, Setting=ZS2026.050.06———
Gemini Embedding 2Year=2026, Setting=ZS2026.050.052———
AVFICSetting=Zero-shot2025.070.03—0.175—
SGD-CLIPTime=347.48s2026.040.02780.11390.2077—
UniConTime=13.45s2026.040.02490.1110.1789—
No alignment2026.0400.00290.0096—