Large-Scale Long-Tailed Recognition in an Open World

We tried all possible methods we could think of (data augmentation, sampling techniques, few-shot learning, imbalanced classification, etc.); but none of the existing methods could handle abundant classes, scarce classes and open classes at the same time (Fig. 1). We define OLTR as learning from long-tail and open-end distributed data and evaluating the classification accuracy over a balanced test set which includes head, tail, and open classes in a continuous spectrum (Fig. 2). Figure 2: Our task of open long-tailed recognition must learn from long-tail distributed training data in an open world and deal with imbalanced classification, few-shot learning, and open-set recognition over the entire spectrum.

Source: bair.berkeley.edu