kevinniechen / face-recognition

Face Recognition with Bayes/KNN using PCA/LDA reduced face vectors

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Face Recognition

Note

Required Code/ folder is split into notebooks/ (exploratory notebooks with plots) and src (simplified classifiers) folders. See below for more information.

Dependencies

  • Python3
  • Run pip install -r requirements to install required Python libaries

Quick Start

Run make run or python3 -W ignore test.py to run classifier models from src/. NOTE: You must be in 'src/' before running test.py or else relative paths will fail. These models contain a fraction of the code in the notebooks.

Run jupyter notebook or jupyter lab (requires Python module installation of jupyter and jupyterlab, respectively) and traverse notebooks/ to run the notebooks. analysis.ipynb contains all exploratory analysis (plots, etc) performed in the report.

Run python3 process_data.py to manually generate the data from the raw inputs from src/. This shouldn't be necessary since the data is already in the .zip file. This code is from the process_data.ipynb notebook.

Directory Structure

notebooks/ contains all exploratory code in Jupyter Notebooks src/ contains refined code i.e classifiers, modules, etc data/ contains raw data in Matlab files, and is populated with images after preprocessing Makefile contains commands to clean data, run models, and perform other helpful tasks report.pdf contains the finalized report reports contains plots used in the report

All code in this repository is written in Python 3.6.

About

Face Recognition with Bayes/KNN using PCA/LDA reduced face vectors


Languages

Language:Jupyter Notebook 98.6%Language:Python 1.4%Language:Makefile 0.0%