Pablo Garcia Moreno — accepted AI-conference papers

2021

Tomographic Auto-Encoder: Unsupervised Bayesian Recovery of Corrupted Data

Francesco Tonolini, Pablo Garcia Moreno, Andreas Damianou, Roderick Murray-Smith

ICLR 2021poster

We propose a new probabilistic method for unsupervised recovery of corrupted data. Given a large ensemble of degraded samples, our method recovers accurate posteriors of clean values, allowing the exploration of the manifold of possible reconstructed data and hence characterising the underlying unce…

Cited by 2 Source ↗PDF ↗Scholar ↗

2020

Empirical Bayes Transductive Meta-Learning with Synthetic Gradients

Shell Xu Hu, Pablo Garcia Moreno, Yang Xiao, Xi Shen, Guillaume Obozinski, Neil Lawrence +1

ICLR 2020poster

We propose a meta-learning approach that learns from multiple tasks in a transductive setting, by leveraging the unlabeled query set in addition to the support set to generate a more powerful model for each task. To develop our framework, we revisit the empirical Bayes formulation for multi-task le…

Cited by 186 Source ↗Scholar ↗