Uncertainty in Model-Agnostic Meta-Learning using Variational Inference

Cuong Nguyen, Thanh-Toan Do, Gustavo Carneiro; Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision (WACV), 2020, pp. 3090-3100

Abstract


We introduce a new, rigorously-formulated Bayesian meta-learning algorithm that learns a probability distribution of model parameter prior for few-shot learning. The proposed algorithm employs a gradient-based variational inference to infer the posterior of model parameters for a new task. Our algorithm can be applied to any model architecture and can be implemented in various machine learning paradigms, including regression and classification. We show that the models trained with our proposed meta-learning algorithm are well calibrated and accurate, with state-of-the-art calibration and classification results on three few-shot classification benchmarks (Omniglot, mini-ImageNet and tiered-ImageNet), and competitive results in a multi-modal task-distribution regression.

Related Material


[pdf] [supp] [video]
[bibtex]
@InProceedings{Nguyen_2020_WACV,
author = {Nguyen, Cuong and Do, Thanh-Toan and Carneiro, Gustavo},
title = {Uncertainty in Model-Agnostic Meta-Learning using Variational Inference},
booktitle = {Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision (WACV)},
month = {March},
year = {2020}
}