Meta-Learning with Differentiable Convex Optimization

Code

• github.com/kjunelee/MetaOptNet
  OfficialIn paperpytorch★ 0
• github.com/learnables/learn2learn
  pytorch★ 2,878
• github.com/cyvius96/few-shot-meta-baseline
  pytorch★ 653
• github.com/yinboc/few-shot-meta-baseline
  pytorch★ 653
• github.com/xiangyu8/PT-MAP-sf
  pytorch★ 24
• github.com/nupurkmr9/S2M2_fewshot
  pytorch★ 0
• github.com/goldblum/AdversarialQuerying
  pytorch★ 0

Abstract

Many meta-learning approaches for few-shot learning rely on simple base learners such as nearest-neighbor classifiers. However, even in the few-shot regime, discriminatively trained linear predictors can offer better generalization. We propose to use these predictors as base learners to learn representations for few-shot learning and show they offer better tradeoffs between feature size and performance across a range of few-shot recognition benchmarks. Our objective is to learn feature embeddings that generalize well under a linear classification rule for novel categories. To efficiently solve the objective, we exploit two properties of linear classifiers: implicit differentiation of the optimality conditions of the convex problem and the dual formulation of the optimization problem. This allows us to use high-dimensional embeddings with improved generalization at a modest increase in computational overhead. Our approach, named MetaOptNet, achieves state-of-the-art performance on miniImageNet, tieredImageNet, CIFAR-FS, and FC100 few-shot learning benchmarks. Our code is available at https://github.com/kjunelee/MetaOptNet.

Tasks

Benchmark Results

Reproductions