MRFDCNet: Multireceptive Field Dense Connection Network for Real-Time Semantic Segmentation

Table 3

Comparison of individual category results on the Cityscapes test set.

Method	Roa	Sid	Bui	Wal	Fen	Pol	TLi	TSi	Veg	Ter
ENet	96.3	74.2	85	32.1	33.2	43.4	34.1	44	88.6	61.4
CGNet	95.9	73.9	89.9	43.9	46	52.9	55.9	63.8	91.7	68.3
EDANet	97.8	80.6	89.5	42	46	52.3	59.8	65	91.4	68.7
ESPNet	92.5	68.5	45.9	89.9	40	47.7	40.7	36.4	54.9	60.3
ESPNetv2	97.3	78.6	88.8	43.5	42.1	49.3	52.6	60.0	90.5	66.8
DABNet	97.8	80.7	90.2	47.9	48.1	56.4	61.8	67	92	69.5
FPENet	96.4	71.7	84.6	27.1	28.8	43.2	39.2	34.4	89.3	61.3
ERFNet	97.7	81	89.8	42.5	48	56.2	59.8	65.3	91.4	68.2
MRFDCNet	98.2	83.7	91.5	50	53.9	59.6	66.6	70.7	92.4	70.4
Method	Sky	Per	Rid	Car	Tru	Bus	Tra	Mot	Bic	mIoU
ENet	90.6	65.5	38.4	90.6	36.9	50.5	48.1	38.8	55.4	58.3
CGNet	94.1	76.7	54.2	91.3	41.3	55.9	32.8	41.1	60.9	64.8
EDANet	93.6	75.7	54.3	92.4	40.9	58.7	56	50.4	64	67.3
ESPNet	92.5	68.5	45.9	89.9	40	47.7	40.7	36.4	54.9	60.3
ESPNetv2	93.3	72.9	53.1	91.8	53.0	65.9	53.2	44.2	59.9	66.2
DABNet	94.3	80.3	59.2	93.7	46	57.1	35	50.4	66.8	68.1
FPENet	92.3	68.1	42.7	89.8	29.1	38.9	27.5	29.1	54.5	55.2
ERFNet	94.2	76.8	57.1	92.8	50.8	60.1	51.8	47.3	61.7	68.0
MRFDCNet	94.7	82	63.0	94.4	55.5	70.6	57.3	59	70.4	72.8