Unlocking Transfer Learning for Open-World Few-Shot Recognition

TL;DR

This paper presents a novel two-stage transfer learning approach with open-set awareness for Few-Shot Open-Set Recognition, significantly improving performance on benchmarks with minimal additional training effort.

Contribution

It introduces a combined open-set aware meta-learning and transfer learning framework, including a strategy to simulate open-set examples, advancing FSOSR capabilities.

Findings

Achieves state-of-the-art results on miniImageNet and tieredImageNet

Requires only 1.5% more training effort than existing methods

Effectively incorporates open-set recognition into transfer learning

Abstract

Few-Shot Open-Set Recognition (FSOSR) targets a critical real-world challenge, aiming to categorize inputs into known categories, termed closed-set classes, while identifying open-set inputs that fall outside these classes. Although transfer learning where a model is tuned to a given few-shot task has become a prominent paradigm in closed-world, we observe that it fails to expand to open-world. To unlock this challenge, we propose a two-stage method which consists of open-set aware meta-learning with open-set free transfer learning. In the open-set aware meta-learning stage, a model is trained to establish a metric space that serves as a beneficial starting point for the subsequent stage. During the open-set free transfer learning stage, the model is further adapted to a specific target task through transfer learning. Additionally, we introduce a strategy to simulate open-set examples by modifying the training dataset or generating pseudo open-set examples. The proposed method achieves state-of-the-art performance on two widely recognized benchmarks, miniImageNet and tieredImageNet, with only a 1.5\% increase in training effort. Our work demonstrates the effectiveness of transfer learning in FSOSR.