Few-shot Object Detection on COCO 2014 (novel) benchmark leaderboard · SOTA2 Research

Few-shot Object Detection on COCO 2014 (novel)

34nAPDE-ViT

Evaluation Results

Method	Links
DE-ViT2023.09		34	29.4	52.9	37
DE-ViT2023.09		34	29.5	53	37.2
DE-ViT2023.09		33.4	28.5	51.4	35.7
DE-ViT2023.09		33.2	28.3	51.4	35.5
DE-ViT2023.09		27.1	24	43.1	28.4
DE-ViT2023.09		26.9	24.2	43.1	28.5
LVC2023.09		26.8	34.8	45.8	27.5
Cross Transformer2023.09		21.4	—	35.5	22.1
NIFF2023.09		20.9	39	—	—
LVC2023.09		19	28.7	34.1	19
NIFF2023.09		18.8	39	—	—
Cross Transformer2023.09		17.1	—	30.2	17
Multi-Relation Det2023.09		16.6	—	31.3	16.1
Meta Faster RCNN2023.09		16.6	—	31.8	15.8
HeteroGraph2023.09		16.5	—	31.9	15.5
FSCE2023.09		16.4	—	—	16.2
DiGeo2023.09		14.2	39.4	26.2	14.8
Retentive RCNN2023.09		13.8	39.3	22.9	13.8
TFA2023.09		13.5	34.5	24.9	13.2
Meta Faster RCNN2023.09		12.7	—	25.7	10.8
FsDetView2023.09		12	9.3	—	—
FSCE2023.09		11.9	—	—	10.5
HeteroGraph2023.09		11.6	—	23.9	9.8
Retentive RCNN2023.09		10.5	39.2	19.5	9.3
DiGeo2023.09		10.3	39.2	18.7	9.9
TFA2023.09		10	33.9	19.2	9.2
Meta R-CNN2023.09		9.9	7.1	25.3	10.8
FSRW2023.09		9.1	—	19	7.6
FsDetView2023.09		7.6	6.4	—	—
Meta R-CNN2023.09		6.1	5.2	19.1	6.6
FSRW2023.09		5.6	—	12.3	4.6