MediaPipe is a Google framework to create pipelines to process data in real time. It was used to be able to capture and draw landmarks on the
human body and hands. Working with OpenCV for the webcam input and video processing, MediaPipe was chosen because it can run on CPU without the need for a GPU.
Learn more about MediaPipe
OpenCV is an open-cource computer vision and machine learning software library.
It is used for video processing and handling webcam input, allowing for real-time video capture and manipulation.
Learn more about OpenCV
The project is written in Python, using libraries and tools for model training,
data handling, and video processing. It was chosen for the simplicity and extensive support for machine learning and computer vision tasks.
It is also compatible with MediaPipem OpenCV, and TensorFlow.
Learn more about Python
TensorFlow makes it easy to build Machine Learning models.
Using the Keras API, a high-level API for building and training models,
to create an LSTM, Long Short-Term Memory model, a type of recurrent neural network (RNN)
made to handle sequential data and learn dependencies over time.
Learn more about TensorFlow
I used two different softwares to be able to use my phone as an external webcam. The integrated one was not clear enough the distance needed away from the camera. So I needed this software to use my phone with out having to deal with the cable length and making sure that I had a stand that wuuld work when my phone was plugged in.