Recognition of Food-Texture Attributes Using an In-Ear Microphone

Food texture is a complex property; various sensory attributes such as perceived crispiness and wetness have been identified ways to quantify it. Objective automatic recognition of these has applications in multiple fields, including health sciences food engineering. In this work we use an in-ear microphone, commonly used for chewing detection, propose algorithms recognizing three food-texture attributes, specifically crispiness, (moisture), chewiness. We binary SVMs, one each attribute, two algorithms: that recognizes attribute at the chew level chewing-bout level. evaluate proposed using leave-one-subject-out cross-validation on dataset with 9 subjects. also them leave-one-food-type-out cross-validation, order examine generalization our approach new, unknown types. Our performs very well (0.95 weighted accuracy new subjects 0.93 types) demonstrates promising results objective