Out of Distribution Detection in Deep Neural Networks – Dr. Arpan Kusari
By: Venu Gopal
Conference Manager, Nxt AI-2026
03/12/2026
We are delighted to welcome Dr. Arpan Kusari, Research Faculty at the University of Michigan Transportation Research Institute (UMTRI), as a featured speaker in the Deep Learning session at Nxt AI-2026.
Dr. Kusari’s research focuses on the robustness and reliability of deep learning systems, particularly for autonomous vehicles (AVs) and advanced driver-assistance systems (ADAS). His work addresses critical challenges in ensuring AI systems remain reliable under unseen and shifting real-world conditions.
In his talk, Dr. Kusari will discuss out-of-distribution (OOD) detection, a key area in modern deep learning that enables models to identify unfamiliar data and improve the safety, robustness, and trustworthiness of AI systems deployed in safety-critical applications.
Why It Matters:
As AI systems become increasingly integrated into transportation, healthcare, and other mission-critical industries, ensuring that models can recognize and respond to unfamiliar situations is essential. OOD detection helps improve reliability and decision-making under uncertain conditions, making AI systems safer and more dependable in real-world deployment scenarios.
About Dr. Arpan Kusari:
- Research Faculty, University of Michigan Transportation Research Institute (UMTRI)
- Expert in deep learning robustness and trustworthy AI
- Researcher in autonomous vehicles (AVs) and ADAS technologies
- Specialist in perception, domain adaptation, and real-world AI deployment