Recognizing Human Actions by a Bag of Visual Words | IEEE Conference Publication | IEEE Xplore