Reuben Brasher

Elmar H. Langholz, Reuben Brasher

We discuss methods for teaching systems to identify gestures such as head nod and shake. We use iPhone X depth camera to gather data and later use similar data as input for a working app. These methods have proved robust for training with limited datasets and thus we make the argument that similar methods could be adapted to learn other human to human non-verbal gestures. We showcase how to augment Euler angle gesture sequences to train models with a relatively large number of parameters such as LSTM and GRU and gain better performance than reported for smaller models such as HMM. In the examples here, we demonstrate how to train such models with Keras and run the resulting models real time on device with CoreML.

Reuben Brasher, Nat Roth, Justin Wagle

We introduce a new metric for measuring how well a model personalizes to a user's specific preferences. We define personalization as a weighting between performance on user specific data and performance on a more general global dataset that represents many different users. This global term serves as a form of regularization that forces us to not overfit to individual users who have small amounts of data. In order to protect user privacy, we add the constraint that we may not centralize or share user data. We also contribute a simple experiment in which we simulate classifying sentiment for users with very distinct vocabularies. This experiment functions as an example of the tension between doing well globally on all users, and doing well on any specific individual user. It also provides a concrete example of how to employ our new metric to help reason about and resolve this tension. We hope this work can help frame and ground future work into personalization.