Prototype Optimization with Neural ODE for Few-Shot Learning
This work addresses prototype bias in few-shot learning, which is a domain-specific problem for machine learning researchers, and is incremental as it builds on existing meta-optimizer frameworks.
The paper tackles prototype bias in few-shot learning by proposing a Neural ODE-based meta-optimizer (MetaNODE) and its efficient extension (E2MetaNODE), achieving superior performance over previous methods and significantly improving computational efficiency without performance degradation.
Few-Shot Learning (FSL) is a challenging task, which aims to recognize novel classes with few examples. Pre-training based methods effectively tackle the problem by pre-training a feature extractor and then performing class prediction via a cosine classifier with mean-based prototypes. Nevertheless, due to the data scarcity, the mean-based prototypes are usually biased. In this paper, we attempt to diminish the prototype bias by regarding it as a prototype optimization problem. To this end, we propose a novel prototype optimization framework to rectify prototypes, i.e., introducing a meta-optimizer to optimize prototypes. Although the existing meta-optimizers can also be adapted to our framework, they all overlook a crucial gradient bias issue, i.e., the mean-based gradient estimation is also biased on sparse data. To address this issue, in this paper, we regard the gradient and its flow as meta-knowledge and then propose a novel Neural Ordinary Differential Equation (ODE)-based meta-optimizer to optimize prototypes, called MetaNODE. Although MetaNODE has shown superior performance, it suffers from a huge computational burden. To further improve its computation efficiency, we conduct a detailed analysis on MetaNODE and then design an effective and efficient MetaNODE extension version (called E2MetaNODE). It consists of two novel modules: E2GradNet and E2Solver, which aim to estimate accurate gradient flows and solve optimal prototypes in an effective and efficient manner, respectively. Extensive experiments show that 1) our methods achieve superior performance over previous FSL methods and 2) our E2MetaNODE significantly improves computation efficiency meanwhile without performance degradation.