Focus areas

Machine Learning, Computer Vision, Computer Graphics, Data Analysis

Key work

• Advanced ML architectures: Transformers, GANs, Diffusion Models, Encoder-Decoders
• 3D geometry processing and video object segmentation
• Built scalable data processing pipelines and real-time systems
• Co-authored research on self-supervised instance segmentation (DINO ViT-Base, achieving +34.15 AP over STEGO baseline on Cityscapes)

Thesis

“AI-based Surrogate Models for Characteristic Curve Prediction” at BMW Group, covering data pipeline, training infrastructure, and production deployment