Microsoft Technology Licensing, LLC

Scott Saponas

Aman Kansal

Daniel Morris

Alice Jane Brush

Julia Maureen Letchner

Desney S Tan

Dan Morris

Ravin Balakrishnan

Jim Turner

Portable devices include environmental sensors that generate metrics about the environment (e.g., accelerometers detecting impulses and vibration, and GPS receivers detecting position and velocity). Such devices often use environmental metrics to extract user input directed at the device by the user, and status information about the device and the environment. Presented herein are techniques for using environmental metrics to infer physical activities performed by the user while attached to the device. For example, jogging may be inferred from regular, strong impulses and typical jogging speed; walking may be inferred from regular, weak impulses and typical walking speed; and riding in a vehicle may be inferred from low-level vibrations and high speed (optionally identifying the type of vehicle ridden by the user). Based on these inferences, the device may automatically present applications and/or or adjust user interfaces suitable for the user's physical activity, rather than responsive to user input.

Physical activity inference from environmental metrics

A charging station wirelessly transmits power to mobile electronic devices (MEDs) each having a planar-shaped receiver coil (RC) and a capacitor connected in parallel across the RC. The station includes a planar charging surface, a number of series-interconnected bank A source coils (SCs), a number of series-interconnected bank B SCs, and electronics for energizing the SCs. Each SC generates a flux field perpendicular to the charging surface. The bank A and bank B SCs are interleaved and alternately energized in a repeating duty cycle. The coils in each bank are also alternately wound in a different direction so that the fields cancel each other out in a far-field environment. Whenever an MED is placed in close proximity to the charging surface, the fields wirelessly induce power in the RC. The MEDs can have any two-dimensional orientation with respect to the charging surface.

Magnetic inductive charging with low far fields

A machine learning model is trained by instructing a user to perform proscribed gestures, sampling signals from EMG sensors arranged arbitrarily on the user's forearm with respect to locations of muscles in the forearm, extracting feature samples from the sampled signals, labeling the feature samples according to the corresponding gestures instructed to be performed, and training the machine learning model with the labeled feature samples. Subsequently, gestures may be recognized using the trained machine learning model by sampling signals from the EMG sensors, extracting from the signals unlabeled feature samples of a same type as those extracted during the training, passing the unlabeled feature samples to the machine learning model, and outputting from the machine learning model indicia of a gesture classified by the machine learning model.

Recognizing gestures from forearm EMG signals

Growing community of inventors

Woodinville, WA, United States of America

Scott Saponas