Lane Detection On Curvelanes

Métriques

F1 score

GFLOPs

Precision

Recall

Résultats

Résultats de performance de divers modèles sur ce benchmark

					Paper Title
CondLSTR (ResNet-101)	88.47	-	-	-	Generating Dynamic Kernels via Transformers for Lane Detection
CondLSTR (ResNet-34)	88.23	-	-	-	Generating Dynamic Kernels via Transformers for Lane Detection
CondLSTR (ResNet-18)	87.99	-	-	-	Generating Dynamic Kernels via Transformers for Lane Detection
CANet-L	87.87	-	91.69	-	CANet: Curved Guide Line Network with Adaptive Decoder for Lane Detection
CANet-M	87.19	22.6	91.53	83.25	CANet: Curved Guide Line Network with Adaptive Decoder for Lane Detection
CANet-S	86.57	13.1	91.37	82.25	CANet: Curved Guide Line Network with Adaptive Decoder for Lane Detection
CLRerNet-DLA34	86.47	18.4	91.66	81.83	CLRerNet: Improving Confidence of Lane Detection with LaneIoU
CondLaneNet-L(ResNet-101)	86.10	44.9	88.98	83.41	CondLaneNet: a Top-to-down Lane Detection Framework Based on Conditional Convolution
CLRNet-DLA34	86.1	18.4	91.4	81.39	CLRerNet: Improving Confidence of Lane Detection with LaneIoU
CondLaneNet-M(ResNet-34)	85.92	19.7	88.29	83.68	CondLaneNet: a Top-to-down Lane Detection Framework Based on Conditional Convolution
CondLaneNet-S(ResNet-18)	85.09	10.3	87.75	82.58	CondLaneNet: a Top-to-down Lane Detection Framework Based on Conditional Convolution
CurveLane-L	82.29	20.7	91.11	75.03	CurveLane-NAS: Unifying Lane-Sensitive Architecture Search and Adaptive Point Blending
CurveLane-M	81.8	11.6	93.49	72.71	CurveLane-NAS: Unifying Lane-Sensitive Architecture Search and Adaptive Point Blending
CurveLane-S	81.12	7.4	93.58	71.59	CurveLane-NAS: Unifying Lane-Sensitive Architecture Search and Adaptive Point Blending
PointLaneNet	78.47	14.8	86.33	72.91	CurveLane-NAS: Unifying Lane-Sensitive Architecture Search and Adaptive Point Blending
SCNN	65.02	328.4	76.13	56.74	CurveLane-NAS: Unifying Lane-Sensitive Architecture Search and Adaptive Point Blending
Enet-SAD	50.31	3.9	63.6	41.6	CurveLane-NAS: Unifying Lane-Sensitive Architecture Search and Adaptive Point Blending
CANet-L(ResNet101)	-	45.7	-	84.36	CANet: Curved Guide Line Network with Adaptive Decoder for Lane Detection

0 of 18 row(s) selected.