Object Detection on COCO-FSOD 30-shot COCO-20 benchmark leaderboard · SOTA2 Research

Object Detection on COCO-FSOD 30-shot COCO-20

45.8nAPFSOD-VFM-L

Evaluation Results

Method	Links
FSOD-VFM-L2026.02		45.8	61.9	49.4
FSOD-VFM-B2026.02		43.7	59.6	47.1
Training-free2025.07		36.8	54.5	38.7
No-Time-To-Train2026.02		36.8	54.5	38.7
CD-VITO (ViT-L)2025.07		35.9	54.5	38
CD-ViTO2024.02		35.9	54.5	38
CD-ViTO (ViT-L)2026.02		35.9	54.5	38
DE-VIT (ViT-L)2025.07		34	52.9	37.2
DE-ViT2024.02		34	52.9	37.2
DE-ViT (ViT-L)2026.02		34	52.9	37.2
CD-ViTO2024.02		33.6	51.6	36.1
DE-ViT2024.02		33.4	51.4	35.7
CD-ViTO2024.02		27.8	43.5	29.5
DE-ViT2024.02		26.9	43.1	28.4
LVC2025.07		26.8	45.8	27.5
LVC2024.02		26.8	45.8	27.5
LVC2026.02		26.8	45.8	27.5
C. Transformer2025.07		21.4	35.5	22.1
Cross-Transformer2024.02		21.4	35.5	22.1
C. Transformer2026.02		21.4	35.5	22.1
NIFF2025.07		20.9	—	—
NIFF2024.02		20.9	—	—
NIFF2026.02		20.9	—	—
Meta F. R-CNN2025.07		16.6	31.8	15.8
Meta Faster R-CNN2024.02		16.6	31.8	15.8
Meta F. R-CNN2026.02		16.6	31.8	15.8
HeteroGraph2025.07		16.5	31.9	15.5
HeteroGraph2024.02		16.5	31.9	15.5
HeteroGraph2026.02		16.5	31.9	15.5
FSCE2025.07		16.4	—	16.2
FSCE2024.02		16.4	—	16.2
FSCE2026.02		16.4	—	16.2
DiGeo2025.07		14.2	26.2	14.8
DiGeo2024.02		14.2	26.2	14.8
DiGeo2026.02		14.2	26.2	14.8
Retentive RCNN2025.07		13.8	22.9	13.8
Retentive RCNN2024.02		13.8	22.9	13.8
Retentive RCNN2026.02		13.8	22.9	13.8
TFA2025.07		13.5	24.9	13.2
TFA2024.02		13.5	24.9	13.2
TFA2026.02		13.5	24.9	13.2
Meta R-CNN2025.07		9.9	25.3	10.8
Meta R-CNN2024.02		9.9	25.3	10.8
Meta R-CNN2026.02		9.9	25.3	10.8
FSRW2025.07		9.1	19	7.6
FSRW2024.02		9.1	19	7.6
FSRW2026.02		9.1	19	7.6