Shahed Masoudian

Shahed Masoudian is a University Assistant at the Institute of Computational Perception, Johannes Kepler University Linz (JKU). His research focuses on machine learning, particularly in transfer learning and domain adaptation. He investigates methods to transfer knowledge from simulated environments to real-world applications, aiming to reduce reliance on large labeled datasets. His work addresses challenges in bias mitigation, audio classification, and neural network optimization.

Key research areas include deep domain adaptation, knowledge distillation, and cognitive biases in recommendation systems. His contributions span applications in acoustic scene classification, industrial condition monitoring, and modular neural network architectures. He has authored/co-authored 15+ peer-reviewed articles since 2022, addressing topics from bias reduction in AI systems to efficient model distillation techniques.

Shahed's educational background includes a master’s thesis supervised by Prof. Gerhard Widmer, exploring simulation-to-real domain adaptation for neural networks. He actively participates in international challenges like DCASE, demonstrating practical solutions for low-complexity audio processing. His research emphasizes bridging the gap between theoretical advancements and real-world deployment, particularly in computationally constrained environments.