List your company
ResearchFormerly WizWand

Scene labeling on SiftFlow (test)

85.3Global AccuracyVGG-conv5-DAG-RNN(8)
69.49273.59677.781.804Sep 2, 2015
Updated 8mo ago

Evaluation results

MethodLinks
VGG-conv5-DAG-RNN(8)Experimental setting=Setting 2, Backbone=VGG-conv5, Neighborhood system=82015.0985.355.7
Long et al.Experimental setting=Setting 22015.0985.251.7
VGG-conv5-ENNExperimental setting=Setting 2, Backbone=VGG-conv52015.098448.8
CNN-65-DAG-RNN(8)Experimental setting=Setting 1, Backbone=CNN-65, Neighborhood system=82015.0981.148.2
CNN-65-DAG-RNN(4)Experimental setting=Setting 1, Backbone=CNN-65, Neighborhood system=42015.0980.542.6
Shuai et al.Experimental setting=Setting 12015.0980.139.7
Yang et al.Experimental setting=Setting 12015.0979.848.7
Sharma et al.Experimental setting=Setting 12015.0979.633.6
Tighe et al.Experimental setting=Setting 12015.0979.239.2
Farabet et al.Experimental setting=Setting 12015.0978.529.4
Pinheiro et al.Experimental setting=Setting 12015.0977.729.8
CNN-65Experimental setting=Setting 1, Backbone=CNN-652015.0976.132.5
CNN-65-ENNExperimental setting=Setting 1, Backbone=CNN-652015.0976.137
Liu et al.Experimental setting=Setting 12015.0974.8—
Byeon et al.Experimental setting=Setting 12015.0970.122.6