3D Object Detection on nuScenes Nighttime (val) benchmark leaderboard · SOTA2 Research

3D Object Detection on nuScenes Nighttime (val)

42.4mAPBEVFusion

Evaluation Results

Method	Links
BEVFusion2022.05		42.4	—	—	—
TransFusion2022.05		41.8	—	—	—
TransFusion-L2022.05		36.2	—	—	—
CenterPoint2022.05		35.4	—	—	—
LightDiff2024.04		17.6	0.774	0.18	1.108
LightDiff2024.04		17	0.69	0.21	0.62
EnlightenGAN2024.04		13.8	0.786	0.193	0.948
Night Image2024.04		13.4	0.787	0.195	0.957
CLIP_LIT2024.04		13.1	0.791	0.199	0.972
EnlightenGAN2024.04		12.8	0.743	0.204	0.68
Night Image2024.04		12.4	0.746	0.205	0.714
CLIP_LIT2024.04		12.1	0.753	0.211	0.739
Zero-DCE++2024.04		8.9	0.826	0.197	1.029
SNR-Aware2024.04		8.9	0.817	0.193	1.088
ShadowDiffusion2024.04		8.2	0.789	0.224	0.718
Zero-DCE++2024.04		7.7	0.78	0.224	0.787
ShadowDiffusion2024.04		7.2	0.851	0.184	1.242
SNR-Aware2024.04		7.2	0.773	0.251	0.742
SCI-difficult2024.04		6.7	0.828	0.187	1.071
SCI2024.04		6.2	0.829	0.181	1.133
URetinex-Net2024.04		5.3	0.831	0.184	1.114
Exposure Diffusion2024.04		4	0.829	0.179	1.092
URetinex-Net2024.04		3.5	0.782	0.243	0.803
Exposure Diffusion2024.04		3.5	0.796	0.244	0.769
SCI2024.04		3.4	0.783	0.235	0.801
SCI-difficult2024.04		3.2	0.764	0.239	0.774