You will get a MediaPipe AI system that tracks poses and gestures in real time
Project details
You'll get a working real-time pose estimation or gesture recognition
system built with MediaPipe and Python — tested, documented, and ready
to run.
I've built this kind of system before. At NUST's research lab, I
developed an AI Yoga Coach that uses MediaPipe to extract joint angles
in real time, classify poses using a Random Forest model, and deliver
live corrective feedback. That same pipeline is the foundation of what
I'll build for you.
What makes my work different:
• I don't just make it work, I make it readable. Every project comes
with clean, commented code and a README you can actually follow.
• I ask the right questions upfront so there are no surprises mid-project.
• I test on real input (webcam or video) before delivering — not just
on my machine in theory.
Whether you need hand tracking, full-body pose detection, or a custom
gesture classifier, I'll scope it honestly and deliver exactly what
was agreed.
system built with MediaPipe and Python — tested, documented, and ready
to run.
I've built this kind of system before. At NUST's research lab, I
developed an AI Yoga Coach that uses MediaPipe to extract joint angles
in real time, classify poses using a Random Forest model, and deliver
live corrective feedback. That same pipeline is the foundation of what
I'll build for you.
What makes my work different:
• I don't just make it work, I make it readable. Every project comes
with clean, commented code and a README you can actually follow.
• I ask the right questions upfront so there are no surprises mid-project.
• I test on real input (webcam or video) before delivering — not just
on my machine in theory.
Whether you need hand tracking, full-body pose detection, or a custom
gesture classifier, I'll scope it honestly and deliver exactly what
was agreed.
Machine Learning Tools
BERT, Keras, MATLAB, NumPy, OpenCV, pandas, Python, Python Scikit-Learn, PyTorch, scikit-learn, TensorFlow, XGBoostWhat's included
| Service Tiers |
Starter
$30
|
Standard
$75
|
Advanced
$150
|
|---|---|---|---|
| Delivery Time | 3 days | 5 days | 7 days |
Number of Revisions | 1 | 2 | 3 |
Number of Model Variations | 1 | 2 | 3 |
Number of Scenarios | 1 | 3 | 5 |
Number of Graphs/Charts | 0 | 1 | 2 |
Model Validation/Testing | - | - | |
Model Documentation | - | ||
Data Source Connectivity | - | - | |
Source Code |
Optional add-ons
You can add these on the next page.
Fast Delivery
+$15 - $40
Additional Revision
+$10
Model Validation/Testing
(+ 1 Day)
+$20
Model Documentation
(+ 1 Day)
+$15Frequently asked questions
About MAHAD UMAR
Computer Vision Developer | MediaPipe, OpenCV & ESP32 Specialist
Islamabad, Pakistan - 12:47 am local time
Recent work includes:
• AI Yoga Pose Coach — MediaPipe + Random Forest for real-time joint-angle extraction, pose classification, and live corrective feedback (NUST Research Lab)
• Autonomous Driving System — YOLOv8 + OpenCV lane detection pipeline with multi-threaded real-time processing
• IoT Health Monitor — ESP32 + ML on custom PCB ($12 cost vs $500 commercial alternatives), 94.2% prediction accuracy
What I can build for you:
• Pose estimation, hand tracking, gesture recognition (MediaPipe)
• Object detection & tracking (YOLOv8, OpenCV)
• ESP32/Arduino IoT prototypes with sensor integration
• C++ / Python algorithms and simulations
I'm a Computer Engineering student at NUST (CGPA 3.45, top student award multiple semesters) with hands-on research and internship experience. I deliver clean, documented code and communicate clearly throughout every project.
Steps for completing your project
After purchasing the project, send requirements so MAHAD UMAR can start the project.
Delivery time starts when MAHAD UMAR receives requirements from you.
MAHAD UMAR works on your project following the steps below.
Revisions may occur after the delivery date.
Understanding Your Requirements
I review your answers, ask any follow-up questions, and confirm the exact scope before writing a single line of code.
Building & Testing the Pipeline
I build the MediaPipe detection pipeline, test it against your input source, and make sure it runs smoothly and accurately.