Sachin Vernekar
Papers
1
Total Citations
19
H-Index
1
About
Sachin Vernekar is a leading researcher in uncertainty quantification for deep learning, with a primary focus on safety-critical computer vision systems. His seminal work, "Calibrating Uncertainties in Object Localization Task" (2018, 19 citations), addresses a fundamental challenge in autonomous driving and surgical robotics: enabling object detection modules to reliably estimate prediction uncertainties. This contribution is pivotal for safe decision-making, as it allows systems to assess the probability of each detected object's accuracy, thereby reducing catastrophic failures in high-stakes environments. Vernekar's research bridges the gap between theoretical calibration methods and practical deployment, earning recognition for advancing trustworthy AI. His work has been influential in shaping how modern perception systems handle ambiguous or noisy sensor data, directly impacting the robustness of autonomous platforms. By pioneering techniques to quantify model confidence in localization tasks, Vernekar has established himself as a key figure in the intersection of probabilistic machine learning and real-world safety applications.
Research Focus
Key Achievements
Top Papers
- 1Calibrating Uncertainties in Object Localization Task19 citations · 2018