Bernardo
Ribeiro.

I’m a PhD candidate at Eindhoven University of Technology’s AIMS Lab. My research focuses on multimodal AI for anomaly detection, with interests in generative models, resource-efficient AI, and computer vision.

GitHub LinkedIn Google Scholar

Background

Master’s in Artificial Intelligence · Cum laude, 2025

Research

Selected publications

All publications

WACV 2027

Generative modeling · Anomaly detection

Flownomaly: Single-Step Conditional-Mean Recovery for Few-Shot Anomaly Detection

B. P. Ribeiro, E. Akdag, E. Bondarev

Conditional flow matching for few-shot visual anomaly detection through single-step recovery in frozen DINOv2 feature space. Evaluated on MVTec-AD, VisA, and Real-IAD.

Paper Code

ICLR 2026

Generative modeling · Weather forecasting

FlowCast: Advancing Precipitation Nowcasting with Conditional Flow Matching

B. P. Ribeiro, J. F. Pucer

Conditional flow matching for probabilistic precipitation nowcasting, with efficient sampling in 1–10 steps. Evaluated on the SEVIR and ARSO benchmarks.