Calvin Galagain

PhD Candidate: Semantic Aware Dynamic SLAM for Embedded Systems.

Robotics

PhD researcher at CEA LIST and ENSTA Paris, focused on real-time semantic SLAM, dynamic-scene perception and efficient neural networks for resource-constrained robotic systems. I work at the boundary between research and engineering: designing perception algorithms, optimizing them for embedded deployment, and integrating them into C++/Python robotics stacks.

PhD Researcher in Embedded Robotic Perception at CEA LIST / ENSTA Paris (2023-10 – Present)

• Developing semantic SLAM methods for dynamic scenes, using semantic and panoptic segmentation to reduce moving-object artifacts and improve map consistency.
• Designing lightweight segmentation pipelines for robotics, with emphasis on predictable latency, memory footprint and deployability on constrained embedded platforms.
• Optimizing neural perception models for edge execution with TensorRT, CUDA-aware deployment, FP16 inference and NVIDIA Jetson evaluation.
• Benchmarking SLAM and segmentation trade-offs across accuracy, runtime and compute budget to guide embedded robotics architecture decisions.

AI Research Intern, 3D Robotic Perception at Exwayz (2023-04 – 2023-10)

• Built real-time object detection and tracking components from 3D LiDAR data for robotic perception stacks.
• Worked on SLAM, SLAM-GPS fusion, anisotropic point-cloud processing and dynamic bounding-box estimation in production-oriented Python/C++ codebases.

AI Research Intern at Zentrum für Recht und Digitalisierung (2022-05 – 2022-09)

• Built real-time video-analysis tools for sensitive-content detection in collaboration with legal experts.
• Explored adversarial learning techniques to assess vulnerabilities in deep neural networks.

Computer Vision Research Project at IRCGN (2021-11 – 2022-09)

• Studied deepfake detection and forensic face recognition from police video footage.
• Implemented landmark detection, data augmentation and deep-learning decoders from scratch.

PhD in AI and Robotics; Semantic Aware Dynamic SLAM for Embedded Systems – CEA LIST / ENSTA Paris (2023)

M2 MVA in Mathematics, Vision and Learning – ENS Paris-Saclay (2022 – 2023)

Engineering Degree in AI, data science, electronics and embedded systems – ESIEE Paris (2020 – 2023)