Research Engineer

• Built a real-time SLAM pipeline combining monocular VO and depth estimation, achieving ~20–25 FPS on noisy indoor data
• Implemented keyframe selection and pose graph optimization, reducing drift to <0.8 m ATE over ~100 m trajectories
• Integrated depth-based reconstruction to improve spatial consistency and reduce scale drift across long sequences
• Debugged sensor noise, calibration drift, and VO failures, improving tracking stability and reducing dropouts by ~30%
• Collected and validated multi-modal datasets, ensuring temporal alignment and robustness for downstream pipelines