Pratik Kalshetti

Senior Lead Engineer @ Qualcomm

About

Researcher and Systems Engineer with 8+ years of experience in 3D Computer Vision and Computer Graphics. Core research on topics including Neural Reconstruction, Gaussian Splatting, Inverse Rendering, Pose Estimation, and SLAM, with first-author papers published at top conferences/journals, including SIGGRAPH and TPAMI, and recognized by multiple research awards and industry fellowships. Systems architect with expertise in mapping 3D reconstruction algorithms to specialized hardware for high-performance applications in AR/VR, Robotics, and Autonomous Driving.

Country

India

City

Bengaluru

Industry

Computer Software

Skill

Hand Tracking, 3D Reconstruction, 3D Computer Vision, Research Skills, Technical Writing, Image Processing, Optimization, Computer Graphics, Machine Learning, Computer Vision

Experience

Senior Lead Engineer

Qualcomm

2024-8 - Present · 2 yrs 2 mos

Bengaluru, Karnataka, India

- Systems architect for mapping 3D Reconstruction algorithms to a hardware accelerator. - Owned and led the systems development of a new algorithm block for an upcoming VR chipset. - Improved the algorithm, leading to a substantial reduction in area and cycle count. - Collaborate with hardware, firmware, and software teams located across various geographies. - Mentored cross-functional teams on computer graphics fundamentals and 3D reconstruction.

Postdoctoral Research Fellow

Max Planck Institute for Informatics

2024-5 - 2024-8 · 4 mos

Collaborator: Dr. Vladislav Golyanik Project: High frame rate tracking of human hands from an event camera. Idea: 3D Gaussian splatting with Real-time Detection Transformers for fast hand reconstruction.

Research Intern

Adobe

2021-7 - 2021-11 · 5 mos

San Francisco Bay Area

Collaborators: Alec Jacobson, Oliver Wang, Thibault Groueix, Vova Kim Project: Efficient video editing with high-level cues. Idea: Propagate frame-level edits via texture registration using proposed mesh optical flow.

Education

Indian Institute of Technology, Bombay

Computer Science

2018-7 - 2024-5 · 5 yrs 11 mos

My Ph.D. thesis focuses on Reconstructing Hand Shape, Pose and Appearance for Tracking from Monocular Video, advised by Parag Chaudhuri. Specifically, my work encompasses personalized shape modeling, articulated pose estimation, 3D model-based tracking using non-linear optimization, and illumination-aware appearance reconstruction using Monte Carlo path tracing in a differentiable deffered rendering framework.