Nassim Belmecheri

Papers

Total Citations

H-Index

About

Nassim Belmecheri is an emerging researcher specializing in software testing methodologies applied to autonomous systems and human-computer interaction. Their work sits at the intersection of machine learning validation and autonomous navigation, with a particular focus on developing robust evaluation frameworks for human trajectory prediction — a critical challenge in the safe deployment of self-driving vehicles and mobile robots. Belmecheri's most notable contribution lies in pioneering the application of metamorphic testing to human trajectory prediction systems. Recognizing that trajectory prediction is inherently stochastic — making traditional oracle-based testing insufficient — their 2024 paper introduced metamorphic testing as a principled approach to evaluating these probabilistic models, garnering 2 citations shortly after publication. This work was extended in 2025 to address multimodal prediction scenarios, reflecting the growing complexity of real-world autonomous system requirements. Though still in the early stages of their research career, Belmecheri is carving out a distinctive niche by addressing a genuinely difficult validation problem: how do you meaningfully test systems whose outputs are inherently uncertain? Their contributions offer valuable tools for researchers and engineers working to build safer, more reliable autonomous systems operating alongside humans.