Abstract: A portable steam generator includes a power source in a housing, a water tank in the housing, a heating element in the housing and configured to heat water in the water tank to produce steam using power from the power source, and a track on the housing configured to support the portable steam generator in an installation position and allow manual removal from the installation position. A pre-plumbed integrated bathroom panel includes a plumbing inlet configured to provide water to the panel, a water outlet connected to the plumbing inlet, a steam outlet connected to the plumbing inlet and a heating element, a sanitizing outlet connected to the plumbing inlet and an additive generator, valves for directing water from the plumbing inlet to the at least one water outlet, the at least one steam outlet, and the at least one sanitizing outlet, and a controller configured to operate at least the plurality of valves.

Abstract: Example solutions provide saliency for anchor-based object detection, and include: performing, with an object detector, a first object detection process on an image, wherein the first object detection process employs a plurality of anchor boxes; identifying an object detection result for the image; determining, from among the plurality of anchor boxes, a first anchor box associated with the object detection result; and while limiting the object detector to the first anchor box, generating, with the object detector, a saliency map for the image. In some examples, the saliency map is used for selecting further training data for the object detector. In some examples, the saliency map comprises a gradient-based saliency map, and is used for auditing or debugging the object detector.

Abstract: Example solutions provide saliency for anchor-based object detection, and include: performing, with an object detector, a first object detection process on an image, wherein the first object detection process employs a plurality of anchor boxes; identifying an object detection result for the image; determining, from among the plurality of anchor boxes, a first anchor box associated with the object detection result; and while limiting the object detector to the first anchor box, generating, with the object detector, a saliency map for the image. In some examples, the saliency map is used for selecting further training data for the object detector. In some examples, the saliency map comprises a gradient-based saliency map, and is used for auditing or debugging the object detector.

Abstract: A respiratory sensor system provides the ability to determine a respiration rate. The respiratory sensor system may include a respiratory sensor and a software solution, where the software solution determines respiration rate from the respiratory sensor (e.g., microphone) located on the bridge of the user's nose. The respiratory sensor may be placed on or near the nose, such as a microphone located in the nosepiece of a pair of glasses. Due to the placement of the microphone and the respiration detection system design, the wearer respiration rate can be determined reliably even with ambient noise and movement. This makes the respiration rate calculation accessible and reliable, both for everyday wear and for extreme situations such as sports.

Abstract: System and techniques for gesture capture are described herein. A first representation of a sensor array may be collected into a first sensor composition. The first sensor composition may be filtered to output a pose status. Here, the pose status corresponds to a pose in a library of poses. A second representation of the sensor array may be collected into a second sensor composition. A fit of the second sensor composition to a model may be measured. The fit may be provided to a gesture consumer based on the fit and the pose.

Abstract: Illustrative examples of a sports apparatus and methods for manufacturing a sports apparatus for sensing a scoring event, including but not limited to hockey pucks. In some examples, the sports apparatus includes a body formed of a substrate material and an additive embedded in the substrate material. The additive may be used by a sensor array to locate the position of the sports apparatus in relation to a sports goal. Example additives include an electronic object or a doping material.

Abstract: System and techniques for gesture capture are described herein. A first representation of a sensor array may be collected into a first sensor composition. The first sensor composition may be filtered to output a pose status. Here, the pose status corresponds to a pose in a library of poses. A second representation of the sensor array may be collected into a second sensor composition. A fit of the second sensor composition to a model may be measured. The fit may be provided to a gesture consumer based on the fit and the pose.

Abstract: Embodiments of the disclosure are directed to systems, methods, and devices for monitoring inertial data from a player swinging a piece of sports equipment. In embodiments, a monitoring device for monitoring swing metrics. The monitoring device includes an inertial measurement unit (IMU) implemented at least partially in hardware to sense inertial data corresponding to a swing of a piece of sports equipment; a processor implemented at least partially in hardware to extrapolate, based on the inertial data, one or more swing parameters associated with the swing of the piece of sports equipment; and a transceiver implemented at least partially in hardware to transmit the one or more swing parameters to a remote device across a wireless connection.

Abstract: A respiratory sensor system provides the ability to determine a respiration rate. The respiratory sensor system may include a respiratory sensor and a software solution, where the software solution determines respiration rate from the respiratory sensor (e.g., microphone) located on the bridge of the user's nose. The respiratory sensor may be placed on or near the nose, such as a microphone located in the nosepiece of a pair of glasses. Due to the placement of the microphone and the respiration detection system design, the wearer respiration rate can be determined reliably even with ambient noise and movement. This makes the respiration rate calculation accessible and reliable, both for everyday wear and for extreme situations such as sports.

Abstract: Illustrative examples of a sports apparatus and methods for manufacturing a sports apparatus for sensing a scoring event, including but not limited to hockey pucks. In some examples, the sports apparatus includes a body formed of a substrate material and an additive embedded in the substrate material. The additive may be used by a sensor array to locate the position of the sports apparatus in relation to a sports goal. Example additives include an electronic object or a doping material.

Abstract: Systems and techniques for a sensor network for trick classification are described herein. A first data stream may be received from a first sensor array affixed to a first free-moving body of a sporting device. A second data stream may be received from a second sensor array affixed to a second free-moving body of the sporting device. A trick region of a predetermined length of the first data stream corresponding with an occurrence of a trick may be determined using data from the first data stream. The trick may be classified using a first set of data from the first data stream corresponding with the trick region and a second set of data from the second data stream. The second set of data may be obtained by aligning the trick region of the first data stream with the second data stream.

Abstract: A system for skydiving trajectory and coordination feedback for a user. The system includes a controller coupled to a plurality of sensor inputs. The controller determines a user drag through air based on input data from the plurality of sensor inputs. The controller also receives a user flightpath and determines a projected user trajectory based on the user drag. The controller also determines user drag adjustments to align the projected user trajectory with the received user flightpath. At least one of an aural, haptic, or visual feedback device is coupled to the controller to provide feedback to the user in response to the user drag adjustments.

Abstract: Systems and techniques for predictive classification in action sports are described herein. A start point for an action may be identified in a data stream including a plurality of data sets corresponding to the action. The data stream may be collected from a sensor array. Action performance features may be extracted from the data stream subsequent to the start point. The action performance features may be compared in real-time to a set of statistical models. A label may be selected for the action based on the comparison. A likelihood of success may be generated for the action based on the comparison. The label for the action and the likelihood of success may be output for display on a display device.

Abstract: Systems and techniques for accelerated pattern matching in action sports are described herein. A plurality of data sets may be collected from a data stream received from a sensor array. Each data set of the plurality of data sets may be compressed into a magnitude value. A first mean and a first standard deviation may be determined for a first set of magnitude values corresponding to a first time period. A second mean and a second standard deviation may be determined for a second set of magnitude values corresponding to a second time period. A mean ratio may be calculated for the first mean and the second mean and a standard deviation ratio may be calculated for the first standard deviation and the second standard deviation. A start point for an action may be generated by determining that the mean ratio and the standard deviation ratio are outside a threshold.

Abstract: A system for skydiving trajectory and coordination feedback for a user. The system includes a controller coupled to a plurality of sensor inputs. The controller determines a user drag through air based on input data from the plurality of sensor inputs. The controller also receives a user flightpath and determines a projected user trajectory based on the user drag. The controller also determines user drag adjustments to align the projected user trajectory with the received user flightpath. At least one of an aural, haptic, or visual feedback device is coupled to the controller to provide feedback to the user in response to the user drag adjustments.

Abstract: Various systems and methods for implementing crowd gesture recognition are described herein. A system for implementing crowd gesture recognition includes an accelerometer; a gyrometer; a gesture detection circuit to: detect an air gesture performed by a user of the system based on data from the accelerometer and gyrometer; and parameterize an intensity of the air gesture; a processor subsystem to determine a transmission frequency band and a transmission strength based on the air gesture and the intensity of the air gesture; and a transducer to transmit a signal on the transmission frequency band with the transmission strength.

Abstract: Systems and techniques for a sensor network for trick classification are described herein. A first data stream may be received from a first sensor array affixed to a first free-moving body of a sporting device. A second data stream may be received from a second sensor array affixed to a second free-moving body of the sporting device. A trick region of a predetermined length of the first data stream corresponding with an occurrence of a trick may be determined using data from the first data stream. The trick may be classified using a first set of data from the first data stream corresponding with the trick region and a second set of data from the second data stream.