Simulated UAV Face Tracking: Python/OpenCV/Webots

I am looking for someone with experience in Webots, Python and computer vision to help develop a simulated UAV/drone camera system for a project.
The aim is to create a simple but fully working Webots simulation where a UAV/drone follows a predefined flight path through an environment containing several human models. The drone camera feed should then be processed in Python to detect and track faces in real time.
The project should include:
• A Webots environment containing a UAV/drone with a working camera
• Several human models positioned within the environment
• A simple predefined UAV flight path or waypoint route
• Access to the simulated drone camera feed using Python
• Real-time face detection using OpenCV YuNet and MediaPipe Face Detection
• Bounding boxes around detected faces
• Basic tracking with persistent IDs so individual faces can be followed between frames
• Live FPS/performance measurement
• Annotated output video showing detections, tracking IDs and FPS
I would also like the system to allow basic testing of different conditions such as:
• Different camera-to-person distances
• Different face/camera angles
• Different UAV movement speeds
• Multiple people in the scene
• Partial occlusion or people crossing each other
Where possible, results should be recorded so the performance of YuNet and MediaPipe can be compared, including detection success, FPS, tracking stability, missed detections and any obvious failures.
Documentation Required
A very important part of the work is a detailed technical report explaining exactly how the entire system was created.
The report must include:
• Step-by-step instructions from initial Webots setup through to the completed system
• Screenshots throughout the development process
• Explanation of how the Webots world, UAV, camera, human models and flight path were configured
• Explanation of how Python communicates with Webots
• Explanation of the face detection and tracking methods used
• Explanation of how YuNet and MediaPipe were implemented and compared
• Explanation of how FPS and other results were measured
• Screenshots and examples of the final simulation working
• Details of any problems/errors encountered during development
• Explanation of how each major problem was diagnosed and solved
• Limitations of the finished system and possible improvements
Files / Handover
I need everything required to reproduce and run the project, including:
• Complete Webots project/world files
• All Python source code
• Models, configuration files and other assets used
• Any required libraries/dependencies with installation instructions
• Face detection model files if required
• Results/data files
• Screenshots
• Annotated demonstration video
• Detailed technical report
• Clear instructions showing exactly how to install and run everything from scratch
The code should be well organised and commented, and I must be able to understand and reproduce the implementation myself.
Budget & Timeframe
Please note that this project has a very limited budget, with a maximum budget of £150. Please do not apply if your quote is significantly above this amount.
The required timeframe is 2 weeks (14 days) from the start of the project. Within this timeframe, I expect the working simulation, source code, project files, results, annotated video, screenshots and detailed technical report to be completed and handed over.
Because of the limited budget and timeframe, I am not looking for an unnecessarily complex simulation. The priority is a simple, reliable and reproducible system that meets all of the requirements above and is properly documented.
Please only apply if you have relevant experience with Webots, Python, OpenCV, MediaPipe and computer vision. When applying, please explain how you would approach the project, what tracking method you would use, confirm that you can complete it within 14 days and the £150 maximum budget, and provide examples of any similar robotics, simulation or computer-vision work you have completed.