Face Detection

Computer Vision | Matlab

This project aims to detect faces in a given image. Face detection was performed by training positive and negative training images, classifying them using support vector machine and detecting faces in the given image using sliding window mechanism.

The following were implemented to identify faces in images.

  • Extracting positive features
  • Extracting random negative features
  • Training positive and negative features using SVM
  • Detecting faces using sliding window
  • Mining hard negatives and retraining SVM
  • Using average face image instead of positive features
  • Using simple Haar-like features




Links: Project Report