PanNote: An Automatic Tool for Panoramic Image Annotation of People’s Positions
Bacchin, A., Barcellona, L., Shamsizadeh, S., Olivastri, E., Pretto, A., Menegatti, E.
IEEE ICRA 2024 · Yokohama · pp. 17006–17012 · doi:10.1109/ICRA57147.2024.10610347
Physical AI · Robotics · Zurich, Switzerland
Physical AI & Robotics Engineer
> working_on:
I build the perception that lets machines understand and act in the real world: embodied perception, multimodal sensor fusion and edge AI, from calibrated sensors and learned models to real-time systems on ROS 2 robots. MSc in AI & Robotics.
$ ros2 topic echo /sepideh/status role: physical_ai_robotics_engineer focus: - physical_ai - robot_perception - sensor_fusion - edge_ai stack: [python, cpp, pytorch, ros2] location: zurich # 47.37°N 8.54°E building: tishtar # phase 2/7 --- $
01 / Results
02 / Experience
$ git log --graph --career
Self-supervised multimodal learning on IMU + GNSS, computer vision for microscopy, and edge–cloud deployment with Docker and Kubernetes. MSCA secondment in Barcelona.
LLM and knowledge-graph initiative in insurance, from concept to prototype.
Real-time YOLOv7 + LiDAR detection, tracking and localisation on mobile robots; camera–LiDAR calibration.
Clinical coding assistant with BiLSTM and attention models, from research to production.
03 / Projects
Active · personal robot
Four wheels, a Raspberry Pi and an STM32, a pan-tilt camera, LiDAR and IMU, all on ROS 2. A platform to take perception, localization and navigation from idea to working hardware.
See the roadmap →
Navigation and obstacle avoidance for a mobile manipulator, with QR-code recognition driving pick-and-place and delivery.
Pick-and-place, conveyor handling and sorting on a physical industrial arm, with joint and Cartesian motion and gripper control.
3D reconstruction pipeline in C++ with SIFT matching, multi-view geometry and bundle adjustment.
[05]Semantic segmentation of LiDAR point clouds on SemanticKITTI, 96% point-wise accuracy on held-out data.
04 / Publication
Bacchin, A., Barcellona, L., Shamsizadeh, S., Olivastri, E., Pretto, A., Menegatti, E.
IEEE ICRA 2024 · Yokohama · pp. 17006–17012 · doi:10.1109/ICRA57147.2024.10610347
05 / Skills
University of Padova, Italy
University of Tehran, Iran