Places-LT

ImagesUnknownIntroduced 2019-01-01

Places-LT has an imbalanced training set with 62,500 images for 365 classes from Places-2. The class frequencies follow a natural power law distribution with a maximum number of 4,980 images per class and a minimum number of 5 images per class. The validation and testing sets are balanced and contain 20 and 100 images per class respectively.

Source: Long-Tailed Recognition Using Class-Balanced Experts

Benchmarks

Few-Shot Image Classification/Top-1 Accuracy Few-Shot Image Classification/Top 1 Accuracy Generalized Few-Shot Classification/Top-1 Accuracy Generalized Few-Shot Classification/Top 1 Accuracy Generalized Few-Shot Learning/Top-1 Accuracy Generalized Few-Shot Learning/Top 1 Accuracy Image Classification/Top-1 Accuracy Image Classification/Top 1 Accuracy Long-tail Learning/Top-1 Accuracy Long-tail Learning/Top 1 Accuracy