MAHDI BAGHERI

Computer Vision and Mechatronic Engineer at Vali Smart Agricultural Systems

Developed artificial falling sorter experimental setup with ~95% accuracy and an MVP achieving ~93% accuracy and ~250 kg/h capacity; optimized YOLO v8 architecture by ~20% with constrained accuracy; investigated knowledge-distillation for context awareness; explored few-shot learning to reduce data labeling.

Computer Vision and Mechatronic Engineer at Vali Smart Agricultural Systems

Developed an artificial falling sorter experimental setup achieving ~95% accuracy and an MVP sorter with ~93% accuracy and ~250 kg/h capacity. Optimized YOLOv8 architecture by ~20% under given constraints and explored knowledge distillation and few-shot learning to improve context awareness and reduce labeling needs.

Computer Vision Research Engineer (Freelancer) at Freelancer Researcher

Built a multi-object tracking system to count cattle at entry/exit in animal husbandry; gathered and labeled datasets; delivered an interactive web app using Dash for user interface.

Freelancer Researcher at Self-employed

Built a multi-object tracking system to count cattle in animal husbandry. Collected, cleaned, and labeled data, and delivered an interactive web app using Dash for user interface.

Computer Vision Research Engineer at ARassis Startup

Developed an automatic face-detection and CT-3D mesh registration system; designed error-assessment for registration and tracking; implemented QR-code-based tracking; created end-to-end calibration for tracking HoloLens 2 with Vive trackers; exploring external tracking enhancements.

Computer Vision Research Engineer at ARassis Startup

Developed an automatic face-detection registration system to align CT-reconstructed meshes with the patient's face, designed end-to-end registration/error assessment, and implemented QR-code-based tracking. Built automatic Hololens 2 calibration with VIVE trackers and contributed to external tracking enhancements.

Technical Product Manager at Agricultural Robotics Lab & Sharif Agrobot

Feasibility study of the product; implemented an experimental setup using Raspberry Pi hardware and YOLOv5-based software; project culminated as the B.Sc. thesis.

Technical Product Manager at Agricultural Robotics Lab & Sharif Agrobot

Conducted feasibility studies for a product and implemented an experimental setup using Raspberry Pi hardware and YOLOv5-based software. The project culminated in the author's B.Sc. thesis.

Computer Vision Intern at CEDRA – Social & Cognitive Robotics Lab

Fine-tuned and trained models for recognizing human facial expressions; created a costume dataset reflecting participants’ expressions.

Computer Vision Intern at Center of Excellence in Design, Robotics, and Automation (CEDRA) – Social & Cognitive Robotics Lab

Fine-tuned and trained a model for recognizing human facial expressions and assembled a dataset reflecting participants’ various expressions.

Researcher (IYPT) at Ghodoosi Highschool

Participated in the Iranian Young Physicists Tournament; contributed to projects such as Circle of Light and Thick Lens; competitions included 2015-2016 timelines.