Liveness Detection with face recognition -- 2

Project Details

This project is about Face "Liveness" and Anti-Spoofing" detection.

It is Not about simple face detection.

I need a skilled strong developer the face detection, recognition.

He must be skilled in Neural network deep learning, opencv, dlib, tensorflow, Machine Learning, python, deep neural network.

Adopting neural network-like deep learning method, non-traditional 3D stereo image sensing technology, can distinguish whether the identified person is a real person, and only needs an HD color camera to identify. I must recognize real and fake faces. This is a Master's and Doctoral Theses in Taiwan.

I want to use deep learning (cnn) to do it. I originally used python to make the following example. However, this example has problems. It is impossible to accurately identify real or fake faces.

[login to view URL]

The code requirements must pass 1. Print attack 2. Replay / video attack. 3D mask attack test to identify real or fake faces, there must be 95% Success rate.

training loss / validation loss / training acc / validation acc After training the model, I can evaluate the results and generate a training graph.

Real-time recognition must be able to detect more than 3 human face frames. The code must have active Heuristic-based algorithms, including eye movement, lip movement,and blink detection. These set of algorithms attempt to track eye movement and blinks to ensure the user is not holding up a photo of another person (since a photo will not blink or move its lips).

Once the job completes, he must provide me demo app, source code. And help employers remotely build the environment and simple process reports.

If you have other ideas you think are better I am happy to discuss.

Thanks in advance.

Python Machine Learning (ML) Algorithm CUDA GPGPU

Project ID: #24502056

About the project