This is a project of pedetrain detection with HOG and SVM. Main steps are: 1.sample image patches -> 2.extract features -> 3.train a SVM model -> 4.collect hard examples -> 5.retrain the model -> 6.demo or test on Test dataset python 2.x or python 3.x Requirements: opencv scikit-learn scikit-image #for demo: cd object-detector python demo.py #train a model cd object-detector python main.py