Machine learning newbie here. Can I put the whole process of face recognition with bag of visual words in a Pipeline? What I mean is, use a Pipeline instead of doing all of