Scene Recognition on MIT Indoor 67
87.1Top-1 AccuracySemantic-Aware Scene Recognition
Evaluation results
| Method | Links | ||
|---|---|---|---|
| Semantic-Aware Scene RecognitionBackbone=RGB Branch* + Sem Branch + G-RGB-H, Number of Parameters=~ 85 M2019.09 | 87.1 | — | |
| SDO (9 scales)Backbone=2xVGG-19, Number of Parameters=~ 276 M2019.09 | 86.76 | — | |
| VSADBackbone=2xVGG-19, Number of Parameters=~ 276 M2019.09 | 86.2 | — | |
| Semantic-Aware Scene RecognitionBackbone=RGB Branch + Sem Branch + G-RGB-H, Number of Parameters=~ 47 M2019.09 | 85.58 | — | |
| RGB BranchBackbone=ResNet-50, Number of Parameters=~ 25 M2019.09 | 84.4 | — | |
| SDO (1 scale)Backbone=2xVGG-19, Number of Parameters=~ 276 M2019.09 | 83.98 | — | |
| RGB BranchBackbone=ResNet-18, Number of Parameters=~ 12 M2019.09 | 82.68 | — | |
| CSBackbone=VGG-19, Number of Parameters=~ 143 M2019.09 | 82.24 | — | |
| CFVBackbone=VGG-19, Number of Parameters=~ 143 M2019.09 | 81 | — | |
| MPP + DSFBackbone=AlexNet, Number of Parameters=~ 62 M2019.09 | 80.78 | — | |
| DSFL + CNNBackbone=AlexNet, Number of Parameters=~ 62 M2019.09 | 76.23 | — | |
| MPP-FCR2 (7 scales)Backbone=AlexNet, Number of Parameters=~ 62 M2019.09 | 75.67 | — | |
| Semantic BranchBackbone=4 Conv, Number of Parameters=~ 2.6 M2019.09 | 73.43 | — | |
| URDL + CNNaugBackbone=AlexNet, Number of Parameters=~ 62 M2019.09 | 71.9 | — | |
| HybridNetBackbone=Places-CNN, Number of Parameters=~ 62 M2019.09 | 70.8 | — | |
| CNNaug-SVMBackbone=OverFeat, Number of Parameters=~ 145 M2019.09 | 69 | — | |
| MOP-CNNBackbone=CaffeNet, Number of Parameters=~ 62 M2019.09 | 68.9 | — | |
| PlaceNetBackbone=Places-CNN, Number of Parameters=~ 62 M2019.09 | 68.24 | — | |
| B-CNNBackbone Model=VGG-VD162019.04 | — | 79 | |
| Deep-TENBackbone Model=ResNet-502019.04 | — | 76.2 | |
| FASONBackbone Model=VGG-VD162019.04 | — | 80.8 | |
| iSQRT-COV-NetBackbone Model=VGG-VD162019.04 | — | 81.4 | |
| iSQRT-COV-NetBackbone Model=ResNet-502019.04 | — | 83.5 | |
| MFAFV-NetBackbone Model=VGG-VD162019.04 | — | 81.1 | |
| SMSOBackbone Model=VGG-VD162019.04 | — | 79.45 | |
| SMSOBackbone Model=ResNet-502019.04 | — | 79.68 | |
| VGG-VD16Backbone Model=VGG-VD162019.04 | — | 67.6 |