Sreenivas Subramoney
Papers
4
Total Citations
40
H-Index
3
About
Sreenivas Subramoney is a leading researcher at the intersection of efficient computing and 3D scene understanding, with a focus on hardware-software co-design for emerging autonomous systems. His work spans visual-inertial odometry (VIO), deep neural network acceleration, and 3D semantic segmentation—critical technologies for AR/VR, robotics, and autonomous driving. Subramoney’s major contributions include pioneering ultra-low-latency, low-power VIO systems for edge devices (19 citations), and developing a unified programmable matrix processor that efficiently handles both deep learning and general matrix algebra workloads (10 citations). He also advanced 3D scene understanding with Segment-Fusion, a hierarchical context fusion method that addresses the persistent part-misclassification problem in semantic segmentation (10 citations), and most recently introduced Ace-of-Spades, an accelerator for spatially sparse 3D convolutions that promises to dramatically improve efficiency for point cloud processing. His work consistently bridges algorithmic innovation with practical hardware implementation, making him a key figure in enabling real-time, energy-efficient perception for autonomous systems.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3
- 4