Mahmut Yurt

Semi-Supervised Learning of Mutually Accelerated Multi-Contrast MRI Synthesis without Fully-Sampled Ground-Truths.

2020

This study proposes a novel semi-supervised learning framework for mutually accelerated multi-contrast MRI synthesis that recovers high-quality images without demanding large training sets of costly fully-sampled source or ground-truth target images. The proposed method presents …

LTDA-Drive: LLMs-guided Generative Models based Long-tail Data Augmentation for Autonomous Driving

2025 · arXiv (Cornell University)

3D perception plays an essential role for improving the safety and performance of autonomous driving. Yet, existing models trained on real-world datasets, which naturally exhibit long-tail distributions, tend to underperform on rare and safety-critical, vulnerable …