Top-1 Error Rate: 67.19
Papers with Code · CC BY-SA 4.0
Weakly-Supervised Object Localization
Weakly supervised object localization (WSOL) learns to localize objects with only image-level labels, no object level labels (bonding boxes, etc.,) is needed. It is more attractive since image-level labels are much easier and cheaper to obtain.
논문
GT-known localization accuracy: 75.0Top-1 Localization Accuracy: 65.2
Top-1 Error Rate: 51.40
GT-known localization accuracy: 65.4Top-1 Localization Accuracy: 52.3
GT-known localization accuracy: 62.45Top-1 Localization Accuracy: 49.39
Top-1 Error Rate: 54.17Top-1 Localization Accuracy: 55.83
GT-known localization accuracy: 68.8Top-1 Localization Accuracy: 56.1average top-1 classification accuracy: 76.7
Top-5 Error: 42.58
Top-1 Error Rate: 37.71MaxBoxAccV2: 66.3
Top-1 Error Rate: 34.8Top-1 Localization Accuracy: 65.22
GT-known localization accuracy: 67.65
MaxBoxAccV2: 90.9
Top-1 Localization Accuracy: 55.58
MaxBoxAccV2: 80.1
Average Top-1 localization accuracy: 59.52
Top-1 Localization Accuracy: 36.00
Top-1 Localization Accuracy: 71.3
Top-1 Localization Accuracy: 74.97