Live Human Detection And Counting Using Tensorflow

What you'll learn
Learn to build a complete human detection model from scratch.
Get to know about Artificial Intelligence, Neural Networks, OpenCV, TensorFlow, and their applications.
Configure the software environment of Anaconda, Jupyter Notebook, and Visual Studio.
Learn to set up python virtual environments and configure pips.
Start by developing code to capture images using the OpenCV library.
Learn about the Image Labelling tool and create annotations.
Get to know about Scripts Records and Label Maps.
Thereafter we will learn about directories creation, defining paths, and their verifications.
We will then understand about TensorFlow Model Garden, WGET Module, and Model API.
Learn and implement protocol buffers and procs.
Get to know about TensorFlow Model Zoo and the usage of pre-trained models.
Learn about Unique IDs, training records, and test record files.
Get to know about Configuration path and writing pipeline configurations and checkpoints.
Learn how to train custom model and evaluate it.
Get to know about the precision, recall, and confusion matrix.
Learn to detect people in the images and videos by using the trained model.
Thereafter, learn to detect people in real time from an external webcam.
After deployment of the model, learn about the freezing graph and saving the final model.
Also, learn the process of converting the human detection model into a TensorFlow lite model.
Finally, learn about archiving the model for editing and building a different model in future.
Requirements
Basic knowledge of Python Programming Language.
Keen to learn and...

Read more

Continue reading...