Abhishek Kar
Papers
1
Total Citations
14
H-Index
1
About
Abhishek Kar is a computer vision researcher whose work lies at the intersection of 3D perception, visual-inertial odometry, and deep learning. His most-cited paper, "Learned Monocular Depth Priors in Visual-Inertial Initialization" (2022, 14 citations), introduces a novel approach that leverages learned depth priors to robustly initialize visual-inertial systems—a critical step for accurate and drift-free state estimation in autonomous navigation and augmented reality. By integrating monocular depth predictions from neural networks into the initialization pipeline, Kar’s work addresses long-standing challenges in scale recovery and sensor fusion, enabling more reliable performance in real-world, texture-poor environments. This contribution bridges the gap between classical geometric methods and modern learning-based techniques, offering a practical solution for systems that rely on cameras and inertial measurement units. Kar’s research is particularly impactful for robotics, drone navigation, and mobile AR applications, where precise, real-time localization is essential. His work exemplifies how learned priors can enhance traditional algorithms, paving the way for more resilient and autonomous visual-inertial systems.
Research Focus
Key Achievements
Top Papers
- 1Learned Monocular Depth Priors in Visual-Inertial Initialization14 citations · 2022