Fabio De Sousa Ribeiro

Papers

Total Citations

H-Index

About

Fabio De Sousa Ribeiro is a researcher whose work sits at the intersection of machine learning, Bayesian inference, and semi-supervised learning. His most cited paper, "Deep Bayesian Self-Training" (2020, 36 citations), introduces a principled framework that combines Bayesian uncertainty estimation with self-training, enabling models to learn more effectively from limited labeled data. This contribution is particularly impactful for domains where annotation is expensive or scarce, such as medical imaging and remote sensing. Ribeiro’s approach leverages uncertainty to guide the pseudo-labeling process, improving robustness and generalization in deep neural networks. Beyond this flagship work, his research explores how probabilistic methods can enhance model calibration and decision-making under uncertainty. With a growing citation record and a focus on bridging theory and practice, Ribeiro is establishing himself as a thoughtful contributor to the next generation of reliable, data-efficient learning systems. His work is especially relevant for students and researchers interested in making deep learning more trustworthy and accessible in real-world, low-resource settings.