Aditya Prakash Patra

Papers

Total Citations

H-Index

About

Aditya Prakash Patra is a researcher at the intersection of computer vision and adversarial machine learning, with a primary focus on the security and robustness of monocular depth estimation systems. His most cited work, "Monocular Depth Estimators: Vulnerabilities and Attacks" (2020), systematically exposes the fragility of modern neural-network-based depth predictors, demonstrating how subtle adversarial perturbations can severely degrade performance—a critical vulnerability for safety-critical applications like autonomous robotics. This paper, with 7 citations, has become a foundational reference for researchers exploring the reliability of visual perception systems. Patra’s contributions highlight a pressing challenge: as monocular depth estimation becomes essential for tasks ranging from navigation to 3D reconstruction, ensuring its resilience against malicious inputs is paramount. His work bridges the gap between cutting-edge depth estimation techniques and practical security concerns, offering both a cautionary analysis and a roadmap for developing more robust models. For students and researchers entering the field, Patra’s research serves as a vital reminder that accuracy alone is insufficient—trustworthy AI must also withstand attack.