List your company

Multi-source sound localization on VGGSound Instruments (test)

89.6CIoU@0.1T-VSL
77.32880.51483.786.886Apr 2, 2024

Evaluation results

MethodLinks
T-VSLEncoders=AudioCLIP (ResNet-50 + ESResNeXt)2024.0489.641.831.5
AVGNEncoders=AudioCLIP pre-trained2024.0486.738.128.8
DSOLEncoders=AudioCLIP pre-trained2024.0485.9—28.4
AlignmentEncoders=AudioCLIP pre-trained2024.0484.935.627.1
Mix-and-LocalizeEncoders=AudioCLIP pre-trained2024.0484.536.326.7
FNACEncoders=AudioCLIP pre-trained2024.0484.135.426.8
AudioCLIPEncoders=AudioCLIP pre-trained2024.0483.834.725.9
MarginNCEEncoders=AudioCLIP pre-trained2024.0482.333.524.7
LVSEncoders=AudioCLIP pre-trained2024.0481.1—23.2
EZ-VSLEncoders=AudioCLIP pre-trained2024.0480.3—22.9
Wav2CLIPEncoders=AudioCLIP pre-trained2024.0480.331.222.2
CoarsetoFineEncoders=AudioCLIP pre-trained2024.0478.8—19.7
OTSEncoders=AudioCLIP pre-trained2024.0477.828.718.2