Table 3 Comparative experimental results on the foggy cityscapes dataset.

Method	mAP50	Person	Rider	Car	Truck	Bus	Train	Motorbike	Bicycle	FPS
Faster RCNN	0.331	0.272	0.336	0.431	0.163	0.258	0.091	0.117	0.269	9.4
SSD300	0.346	0.334	0.379	0.425	0.284	0.366	0.378	0.255	0.314	19
RetinaNet	0.315	0.291	0.397	0.429	0.208	0.374	0.241	0.265	0.299	11
CenterNet	0.242	0.275	0.366	0.377	0.131	0.286	0.027	0.178	0.294	23
DETR	0.276	0.248	0.307	0.415	0.236	0.338	0.197	0.213	0.265	26
EfficentDet	0.264	0.279	0.362	0.352	0.16	0.283	0.102	0.246	0.325	29
Rtmdet	0.338	0.336	0.379	0.485	0.265	0.387	0.236	0.28	0.336	27
YOLOv5-L	0.343	0.299	0.433	0.435	0.235	0.36	0.328	0.301	0.352	45
YOLOv5-X	0.41	0.432	0.478	0.586	0.238	0.457	0.392	0.315	0.382	30
YOLOX-Tiny	0.401	0.399	0.473	0.513	0.279	0.411	0.352	0.363	0.418	45
YOLOv7-s	0.379	0.471	0.497	0.539	0.252	0.389	0.159	0.311	0.414	47
YOLOv8-n	0.398	0.332	0.475	0.479	0.316	0.474	0.409	0.323	0.371	110
YOLOv8-s	0.442	0.44	0.439	0.603	0.316	0.504	0.515	0.317	0.406	81
YOLOv10-n	0.399	0.389	0.499	0.543	0.259	0.483	0.337	0.286	0.396	91
YOLOv11-n	0.408	0.358	0.408	0.595	0.332	0.493	0.427	0.274	0.377	128
MSFFA-YOLO	0.468	0.448	0.433	0.642	0.398	0.551	0.501	0.383	0.388	108
R-YOLO	0.489	0.473	0.495	0.666	0.391	0.558	0.522	0.409	0.444	98
HR-YOLO	0.495	0.45	0.516	0.609	0.424	0.571	0.515	0.417	0.456	108

Quick links

Search