Mohammed Oualid Attaoui
Papers
1
Total Citations
2
H-Index
1
About
Mohammed Oualid Attaoui is a leading researcher at the intersection of deep learning verification, safety-critical systems, and simulation-based testing. His most prominent work introduces a novel, search-based methodology for testing and retraining Deep Neural Networks (DNNs) using Generative Adversarial Network (GAN)-enhanced simulations. This approach directly addresses the critical challenge of validating DNNs in high-stakes domains like autonomous vehicles and robotics, where semantic segmentation errors can have catastrophic consequences. By automating the generation of realistic, adversarial test scenarios, Attaoui’s framework enables the systematic identification of model weaknesses and facilitates targeted retraining, significantly improving robustness. His contributions are foundational for advancing the reliability and trustworthiness of AI in safety-critical applications. With his 2025 paper already garnering early citations, his work is rapidly gaining recognition for providing a practical, scalable solution to the "black box" problem of DNN verification. Attaoui’s research is essential reading for anyone working on dependable AI, autonomous systems, or automated software testing.
Research Focus
Key Achievements
Top Papers
- 1Search-Based DNN Testing and Retraining With GAN-Enhanced Simulations2 citations · 2025