Releases: tyler-cromwell/Retina
v0.5
v0.4
Added class to represent configuration.
Added new tool, compare_faces.py.
Can recognize faces from an image file.
Saves algorithm name in recognizer model filename.
Recognition returns confidence values separately.
Uses -1 for confidence value of an unknown person.
Made Tkinter optional.
v0.3
Added Haar and LBP classifiers
Added settings for dedicated testing machine
Added a script to process raw images for training
Can preview preprocessing steps in display window
Improved preprocessing (recognition improved)
Retrieving confidence value
More bug fixes
v0.2
Data set is created with and without glasses.
Added frame preprocessing.
Added install instructions.
Added dependency list.
Fixed various bugs.
v0.1
Capable of recognizing specific faces once trained.
Added a tool for creating a recognizer training set.
Added a tool for training and saving a recognizer.