10-shot image generation on nuScenes

Metric: IoU veh - 448x800 - Vis filter. - 100x100 at 0.5 (higher is better)

LeaderboardDataset

Loading chart...

Results

Sort:

#	Model↕	IoU veh - 448x800 - Vis filter. - 100x100 at 0.5▼	Extra Data	Paper	Date↕	Code
1	PointBeV	48.7	No	PointBeV: A Sparse Approach to BeV Predictions	2023-12-01	Code
2	PointBeV (static)	47.6	No	PointBeV: A Sparse Approach to BeV Predictions	2023-12-01	Code
3	Simple-BEV	46.6	No	Simple-BEV: What Really Matters for Multi-Sensor...	2022-06-16	Code
4	BEVFormer	45.5	No	BEVFormer: Learning Bird's-Eye-View Representati...	2022-03-31	Code
5	BAEFormer	41	No	-	-	-
6	CVT	37.7	No	Cross-view Transformers for real-time Map-view S...	2022-05-05	Code

#1PointBeVSOTA
48.7
IoU veh - 448x800 - Vis filter. - 100x100 at 0.5· 2023-12-01
PointBeV: A Sparse Approach to BeV Predictions Code
#2PointBeV (static)
47.6
IoU veh - 448x800 - Vis filter. - 100x100 at 0.5· 2023-12-01
PointBeV: A Sparse Approach to BeV Predictions Code
#3Simple-BEVSOTA
46.6
IoU veh - 448x800 - Vis filter. - 100x100 at 0.5· 2022-06-16
Simple-BEV: What Really Matters for Multi-Sensor BEV Perception?Code
#4BEVFormerSOTA
45.5
IoU veh - 448x800 - Vis filter. - 100x100 at 0.5· 2022-03-31
BEVFormer: Learning Bird's-Eye-View Representation from Multi-Camera Images via Spatiotemporal Transformers Code
#5BAEFormer
41
IoU veh - 448x800 - Vis filter. - 100x100 at 0.5
No paper
#6CVT
37.7
IoU veh - 448x800 - Vis filter. - 100x100 at 0.5· 2022-05-05
Cross-view Transformers for real-time Map-view Semantic Segmentation Code