← Search

Daniel Coelho de Castro

5 accepted papers

2024

RadEdit: stress-testing biomedical vision models via diffusion image editing

ECCV 2024poster

"Biomedical imaging datasets are often small and biased, meaning that real-world performance of predictive models can be substantially lower than expected from internal testing. This work proposes using generative image editing to simulate dataset shifts and diagnose failure modes of biomedical visi…

Cited by 9SourcePDFScholar
2020

Deep Structural Causal Models for Tractable Counterfactual Inference

NeurIPS 2020poster

We formulate a general framework for building structural causal models (SCMs) with deep learning components. The proposed approach employs normalising flows and variational inference to enable tractable inference of exogenous noise variables - a crucial step for counterfactual inference that is miss…

2020

Stochastic Segmentation Networks: Modelling Spatially Correlated Aleatoric Uncertainty

NeurIPS 2020poster

In image segmentation, there is often more than one plausible solution for a given input. In medical imaging, for example, experts will often disagree about the exact location of object boundaries. Estimating this inherent uncertainty and predicting multiple plausible hypotheses is of great interest…

2019

Domain Generalization via Model-Agnostic Learning of Semantic Features

NeurIPS 2019poster

Generalization capability to unseen domains is crucial for machine learning models when deploying to real-world conditions. We investigate the challenging problem of domain generalization, i.e., training a model on multi-domain source data such that it can directly generalize to target domains with…

2018

From Face Recognition to Models of Identity: A Bayesian Approach to Learning about Unknown Identities from Unsupervised Data

ECCV 2018poster

Current face recognition systems robustly recognize identities across a wide variety of imaging conditions. In these systems recognition is performed via classification into known identities obtained from supervised identity annotations. There are two problems with this current paradigm: (1) current…

Cited by 10SourcePDFScholar