抄録
In this paper, we proposed a method to improve food image recognition accuracy by extracting better deep features with the latest deep learning network considering mealtime and distributed representation. We also evaluated these features by a personalized classifier.
