Requirements:

1. python 2.7.
2. Liblinear.
3. PIL.

This program trains a logistic regression image classifier. Simply put your data in the data/train Directory as specified and type:

python training.py

To test the classier. put new test images in the data/test dir and type:

python classify.py.

The training consists of:

1. extracting dense SIFT features from all training images.
2. Saving data to file.
3. Feeding the Logistic regression model with data.
4. Saving the model.

Testing consists of:

1. Extracting dense SIFT from a test image.
2. testing the saved model with the test image.
3. Output : predicted labels.