I will build real time pose estimation and action recognition system using mediapipe


About this gig
I will build a real-time pose estimation and action recognition system using Computer Vision, MediaPipe, OpenCV, and Deep Learning. The system can detect human body movements, track poses, recognize actions/exercises, and perform repetition counting from webcam or video input. Suitable for fitness AI, motion analysis, sports analytics, surveillance, and smart vision applications.
Get to know Rehman A.
Computer Vision Engineer
- FromPakistan
- Member sinceOct 2025
Languages
English
My Portfolio
FAQ
What technologies do you use?
I mainly use Python, OpenCV, MediaPipe, TensorFlow/Keras, and Deep Learning models such as GRU/LSTM for real-time pose estimation and action recognition systems.
Can the system work in real-time using webcam?
Yes, I can develop systems that work in real-time using webcam or live video streams depending on the project requirements.
Can you recognize custom exercises or actions?
Yes, custom action or exercise recognition models can be developed and trained using your dataset or publicly available datasets.
Can you integrate the AI model into my application?
Yes, integration into desktop, web, or existing AI applications can be provided in basic package but price might be vary based on complexity of software in which we integrate the AI.
Do you provide deployment support?
Yes, deployment assistance can be provided depending on the platform and project requirements.

