List your company
ResearchFormerly WizWand

Object Counting on FSC-147 (test)

5.03MAEABACUS-3B
4.64647.23579.82512.4143Jan 22, 2022Oct 19, 2022Jul 16, 2023Apr 12, 2024Jan 7, 2025Oct 4, 2025Jul 2, 2026
Updated 4mo ago

Evaluation results

MethodLinks
ABACUS-3BSupervision=Zero-shot/MLLM (Z)2026.065.0327.03β€”β€”β€”β€”β€”β€”β€”
CountGD2026.025.7424.09β€”β€”β€”β€”β€”β€”β€”
CountGDScheme=Few-shot, Exemplar=Visual & Text, Backbone=GDINO-Swin-B2026.036.7543.65β€”β€”β€”β€”β€”β€”β€”
CoDiProtocol=One-shot, Category=Localization-based, Text-prompts=false2025.127.0430.23β€”β€”β€”β€”β€”β€”β€”
CountSEScheme=Few-shot, Exemplar=Text, Backbone=GDINO-Swin-B2026.037.8482.99β€”β€”β€”β€”β€”β€”β€”
CountSESupervision=Point-level (P), visual exemplars=few-shot2026.067.8482.99β€”β€”β€”β€”β€”β€”β€”
GeCo2024.097.9154.28β€”β€”β€”β€”β€”β€”β€”
GeComode=one-shot2024.098.160.16β€”β€”β€”β€”β€”β€”β€”
GeCoProtocol=One-shot, Category=Detection-based, Text-prompts=false2025.128.160.16β€”β€”β€”β€”β€”β€”β€”
CountGD++Supervision=Point-level (P)2026.068.3927.03β€”β€”β€”β€”β€”β€”β€”
CACVITBackbone=ViT-B, Resolution=384, Shots=12023.058.6229.92β€”β€”β€”β€”β€”β€”β€”
CountEx2026.028.6384.55β€”β€”β€”β€”β€”β€”β€”
DAVEFew-shot=true2024.048.6632.36β€”β€”β€”β€”β€”β€”β€”
DAVE2026.028.6632.36β€”β€”β€”β€”β€”β€”β€”
SAVEPrior-Free=true2026.018.9280.39β€”β€”β€”β€”β€”β€”β€”
CACVITBackbone=ViT-B, Resolution=384, Shots=32023.059.1348.96β€”β€”β€”β€”β€”β€”β€”
CACViT2026.029.1348.96β€”β€”β€”β€”β€”β€”β€”
SSDBackbone=ResNet-50, Shot count=3-shot2024.059.5834.13β€”β€”β€”β€”β€”β€”β€”
DCountType=Detection, Backbone=GDino, Decoder=DefTransformer2026.069.83105.15β€”β€”β€”β€”β€”β€”β€”
TrueCount𝑑π‘₯𝑑Type=Detection, Backbone=GDino, Decoder=DefTransformer2026.0610.11106.11β€”β€”β€”β€”β€”β€”β€”
GroundingRECSupervision=point-level2026.0210.12107.19β€”β€”β€”β€”β€”β€”β€”
GroundingRECTrained Model=true2024.1210.12107.19β€”β€”β€”β€”β€”β€”β€”
GroundingREC2026.0210.12107.19β€”β€”β€”β€”β€”β€”β€”
GroundingRECType=Text, Base model=Grounding DINO, Differentiable=Γ—2026.0610.12107.19β€”β€”β€”β€”β€”β€”β€”
GroundingRECType=Detection, Backbone=GDino, Decoder=DefTransformer2026.0610.12107.19β€”β€”β€”β€”β€”β€”β€”
CountDiffProtocol=One-shot, Category=Density-based, Text-prompts=true2025.1210.2957.27β€”β€”β€”β€”β€”β€”β€”
CAD-GD2026.0210.3586.88β€”β€”β€”β€”β€”β€”β€”
CAD-GDType=Text, Base model=Grounding DINO, Differentiable=Γ—2026.0610.3586.88β€”β€”β€”β€”β€”β€”β€”
CAD-GDSupervision=Point-level (P)2026.0610.3586.88β€”β€”β€”β€”β€”β€”β€”
DAVEboxCounting mode=detection-based2024.0410.4574.51β€”β€”β€”β€”β€”β€”β€”
DAVE2024.0910.4574.51β€”β€”β€”β€”β€”β€”β€”
LOCAScenario=few-shot2022.1110.7956.97β€”β€”β€”β€”β€”β€”β€”
LOCA2023.0710.7956.97β€”β€”β€”β€”β€”β€”β€”
LOCAFew-shot=true2024.0410.7956.97β€”β€”β€”β€”β€”β€”β€”
LOCABackbone=ResNet-50, Shot count=3-shot2024.0510.7956.97β€”β€”β€”β€”β€”β€”β€”
LOCA2024.0910.7956.97β€”β€”β€”β€”β€”β€”β€”
LOCATraining-free=false, Prior-free=false, Exemplar-based=true2026.0110.7956.97β€”0.267.96787878β€”
LOCA2026.0210.7956.97β€”β€”β€”β€”β€”β€”β€”
LOCAScheme=Few-shot, Shot=32025.0210.9756.97β€”β€”β€”β€”β€”β€”β€”
LOCAScheme=Few-shot, Exemplar=Visual, Backbone=ResNet-502026.0310.9756.97β€”β€”β€”β€”β€”β€”β€”
DAVEmode=one-shot2024.0911.5486.62β€”β€”β€”β€”β€”β€”β€”
DAVEProtocol=One-shot, Category=Detection-based, Text-prompts=false2025.1211.5486.62β€”β€”β€”β€”β€”β€”β€”
SSDBackbone=ResNet-50, Shot count=1-shot2024.0511.6171.55β€”β€”β€”β€”β€”β€”β€”
TMRScheme=Few-shot, Exemplar=Visual, Backbone=SAM-ViT-H/162026.0311.6357.46β€”β€”β€”β€”β€”β€”β€”
T2ICountScheme=Zero-shot, Shot=02025.0211.7697.86β€”β€”β€”β€”β€”β€”β€”
T2ICountVenue=CVPR’25, inference_only=false2025.1211.7697.86β€”β€”β€”β€”β€”β€”β€”
T2ICountSupervision=point-level2026.0211.7697.86β€”β€”β€”β€”β€”β€”β€”
T2ICountScheme=Zero-shot, Exemplar=Text, Backbone=Stable Diff-v1.52026.0311.7697.86β€”β€”β€”β€”β€”β€”β€”
T2ICountSupervision=Point-level (P)2026.0611.7697.86β€”β€”β€”β€”β€”β€”β€”
T2ICountVenue=CVPR’25, Training=true2026.0711.7697.86β€”β€”β€”β€”β€”β€”β€”
CounTRYear=2022, Backbone=ViT, # Shots=32022.0811.9591.23β€”β€”β€”β€”β€”β€”β€”
CounTRScenario=few-shot2022.1111.9591.23β€”β€”β€”β€”β€”β€”β€”
CounTRBackbone=ViT-B & CNN, Resolution=384, Shots=32023.0511.9591.23β€”β€”β€”β€”β€”β€”β€”
CounTR2023.0711.9591.23β€”β€”β€”β€”β€”β€”β€”
CounTRFew-shot=true2024.0411.9591.23β€”β€”β€”β€”β€”β€”β€”
CounTRBackbone=ViT/ConvNet, Shot count=3-shot2024.0511.9591.23β€”β€”β€”β€”β€”β€”β€”
CounTR2024.0911.9591.23β€”β€”β€”β€”β€”β€”β€”
CounTRTraining-free=false, Prior-free=false, Exemplar-based=true2026.0111.9591.23β€”0.237.44797878β€”
CounTR2026.0211.9591.23β€”β€”β€”β€”β€”β€”β€”
CountTRScheme=Few-shot, Exemplar=Visual, Backbone=ViT-B/162026.0311.9591.23β€”β€”β€”β€”β€”β€”β€”
CounTREvaluation Protocol=one-shot2022.1112.0690.01β€”β€”β€”β€”β€”β€”β€”
CounTRBackbone=ViT-B & CNN, Resolution=384, Shots=12023.0512.0690.01β€”β€”β€”β€”β€”β€”β€”
CounTRBackbone=ViT/ConvNet, Shot count=1-shot2024.0512.0690.01β€”β€”β€”β€”β€”β€”β€”
CounTRmode=one-shot2024.0912.0690.01β€”β€”β€”β€”β€”β€”β€”
CounTRProtocol=One-shot, Category=Density-based, Text-prompts=false2025.1212.0690.01β€”β€”β€”β€”β€”β€”β€”
MambaCountType=Point, Backbone=YOLOE, Decoder=Mamba2026.0612.23105.69β€”β€”β€”β€”β€”β€”β€”
A-Simple-ButPrompting Info=3-shot2026.0112.2656.33β€”β€”β€”β€”β€”β€”β€”
QICA†Scheme=Zero-shot, Exemplar=Text, Backbone=ViT-L/142026.0312.4197.28β€”β€”β€”β€”β€”β€”β€”
QICAVenue=CVPR’26, Training=true2026.0712.4197.28β€”β€”β€”β€”β€”β€”β€”
LOCAEvaluation Protocol=one-shot2022.1112.5375.32β€”β€”β€”β€”β€”β€”β€”
LOCABackbone=ResNet-50, Shot count=1-shot2024.0512.5375.32β€”β€”β€”β€”β€”β€”β€”
LOCAmode=one-shot2024.0912.5375.32β€”β€”β€”β€”β€”β€”β€”
LOCAProtocol=One-shot, Category=Density-based, Text-prompts=false2025.1212.5375.32β€”β€”β€”β€”β€”β€”β€”
CountGDType=Text, Base model=Grounding DINO, Differentiable=Γ—2026.0612.9898.35β€”β€”β€”β€”β€”β€”β€”
COUNTGDType=Detection, Backbone=GDino, Decoder=DefTransformer2026.0612.9898.35β€”β€”β€”β€”β€”β€”β€”
PSECO2024.0913.05112.86β€”β€”β€”β€”β€”β€”β€”
PseCoScheme=Few-shot, Shot=32025.0213.05112.86β€”β€”β€”β€”β€”β€”β€”
PseCoScheme=Few-shot, Exemplar=Visual, Backbone=SAM-ViT-H/162026.0313.05112.89β€”β€”β€”β€”β€”β€”β€”
QICAScheme=Zero-shot, Exemplar=Text, Backbone=ViT-B/162026.0313.05104.17β€”β€”β€”β€”β€”β€”β€”
MAFEAPrior-Free=true2026.0113.23105.99β€”β€”β€”β€”β€”β€”β€”
GeCoScheme=Zero-shot, Shot=02025.0213.3108.72β€”β€”β€”β€”β€”β€”β€”
GeCoPrior-Free=true2026.0113.3108.72β€”β€”β€”β€”β€”β€”β€”
GeCoVenue=NeurIPS’24, inference_only=false2025.1213.3108.72β€”β€”β€”β€”β€”β€”β€”
GeCoScheme=Zero-shot, Exemplar=Visual, Backbone=SAM-ViT-H/162026.0313.3108.72β€”β€”β€”β€”β€”β€”β€”
RT-CounterType=Text, Base model=YOLOE, Differentiable=βœ“2026.0613.3104.63β€”β€”β€”β€”β€”β€”β€”
GeCoVenue=NeurIPS’24, Training=true2026.0713.3108.72β€”β€”β€”β€”β€”β€”β€”
SPDCNBackbone=ResNet-18, Resolution=576, Shots=32023.0513.5196.8β€”β€”β€”β€”β€”β€”β€”
SPDCNBackbone=VGG-19, Shot count=3-shot2024.0513.5196.8β€”β€”β€”β€”β€”β€”β€”
WS-COCSupervision=image-level2026.0213.9197.28β€”β€”β€”β€”β€”β€”β€”
WS-COC-7BSupervision=Image-level (I)2026.0613.9197.28β€”β€”β€”β€”β€”β€”β€”
GCA-SUNPrior-Free=true2026.011492.19β€”β€”β€”β€”β€”β€”β€”
GCA-SUNetCategory=EFC, Backbone=SwinT-B, Resolution=3842024.0914.0884.77β€”β€”β€”β€”β€”β€”β€”
CounTRMode=Zero-shot2022.1114.12108.01β€”β€”β€”β€”β€”β€”β€”
CounTRsetup=zero-shot2024.0414.12108.01β€”β€”β€”β€”β€”β€”β€”
SAFECounttest-time adaptation=false2022.0114.3285.54β€”β€”β€”β€”β€”β€”β€”
SAFECountScenario=few-shot2022.1114.3285.54β€”β€”β€”β€”β€”β€”β€”
SAFECountBackbone=ResNet-18, Resolution=512, Shots=32023.0514.3285.54β€”β€”β€”β€”β€”β€”β€”
SAFECountFew-shot=true2024.0414.3285.54β€”β€”β€”β€”β€”β€”β€”
SAFECountBackbone=ResNet-18, Shot count=3-shot2024.0514.3285.54β€”β€”β€”β€”β€”β€”β€”
BMNet+Fine-tuning=false, Backbone=SwAV2022.0314.62β€”91.83β€”β€”β€”β€”β€”β€”