List your company

Semantic Segmentation Boundary Alignment on Cityscapes (val)

92.3RoadDeepLabV3+
83.35685.6788890.322Jul 12, 2019

Evaluation results

MethodLinks
DeepLabV3+Threshold=12px2019.0792.380.487.259.653.783.875.281.290.260.890.476.678.791.68187.192.681.87880.1——
GSCNNThreshold=12px2019.0792.281.787.959.654.387.182.384.490.961.191.980.482.892.678.59094.679.182.281.8——
GSCNNThreshold=9px2019.0791.380.18658.552.986.181.583.38959.891.179.181.591.578.189.794.478.580.480.7——
DeepLabV3+Threshold=9px2019.0791.278.384.858.152.482.173.779.587.959.489.574.776.89080.586.692.58175.478.7——
GSCNNThreshold=5px2019.0788.775.380.955.949.983.678.680.483.456.688.475.477.888.37788.994.276.975.177.6——
DeepLabV3+Threshold=5px2019.0788.172.678.15549.177.96974.78155.886.46971.985.479.485.492.179.468.474.7——
GSCNNThreshold=3px2019.078568.874.153.34779.674.376.275.353.183.569.873.183.475.88893.975.168.573.6——
DeepLabV3+Threshold=3px2019.0783.765.169.752.246.27262.867.771.85280.961.566.478.878.283.991.777.960.969.7——
CASENetTest NMS=false, Backbone=ResNet101, Pre-trained=ImageNet2020.04————————————————————68.9257.65
CASENet*Test NMS=false, Backbone=ResNet101, Pre-trained=ImageNet2020.04————————————————————69.2961.3
CASENet*Test NMS=true, Backbone=ResNet101, Pre-trained=ImageNet2020.04————————————————————70.3168.13
RPCNetTest NMS=true, Backbone=ResNet101, Pre-trained=ImageNet2020.04————————————————————78.2278.79
STEALTest NMS=false, Backbone=ResNet101, Pre-trained=ImageNet2020.04————————————————————70.6766.44
STEALTest NMS=true, Backbone=ResNet101, Pre-trained=ImageNet2020.04————————————————————71.4268.89