Abstract: Techniques and apparatuses are described that train machine-learned modules to perform radar-based gesture detection in an ambient compute environment. Compared to other smart devices that rely on a physical user interface, a smart device (104) with a radar system (102) can support ambient computing by providing an eye-free interaction and less cognitively demanding gesture-based user interface. The radar system (102) can be designed to address a variety of challenges associated with ambient computing, including power consumption, environmental variations, background noise, size, and user privacy. The radar system (102) uses an ambient-computing machine-learned module (222) to quickly recognize gestures performed by a user up to at least two meters away. The ambient-computing machine-learned module (222) is trained, at least in part, using a two-phase evaluation process, which includes a segmented classification task and an unsegmented recognition task.

Abstract: Techniques and devices for radar-based gesture determination at long ranges are described in this document. The techniques described herein enable a computing device to detect and recognize gestures at long-range extents of up to eight meters. The computing device of this disclosure does not require the user to perform a gestural command at a specific location, in a specific orientation, contingent upon a wake-up trigger, or at a specific time, enabling the user to freely provide commands whenever and wherever is most convenient. This continual recognition of gestures may be enabled by a machine-learned model, generation of augmented data, and inclusion of negative data.

Abstract: Various arrangements for performing contactless fetal movement tracking are detailed herein. User input can first be received from an expectant mother requesting that contactless fetal movement monitoring be performed. A state analysis on the expectant mother may be performed to determine that the expectant mother is present and static in a bed at which the contactless fetal movement tracking device is pointed. Fetal movement tracking may be performed using radar data received from a radar sensor of the contactless fetal movement tracking device while the state analysis indicates that the expectant mother is present and static in the bed. A fetal tracking report may then be presented based on the performed fetal movement tracking.

Abstract: Techniques and apparatuses are described that train machine-learned modules to perform radar-based gesture detection in an ambient compute environment. Compared to other smart devices that rely on a physical user interface, a smart device (104) with a radar system (102) can support ambient computing by providing an eye-free interaction and less cognitively demanding gesture-based user interface. The radar system (102) can be designed to address a variety of challenges associated with ambient computing, including power consumption, environmental variations, background noise, size, and user privacy. The radar system (102) uses an ambient-computing machine-learned module (222) to quickly recognize gestures performed by a user up to at least two meters away. The ambient-computing machine-learned module (222) is trained, at least in part, using a two-phase evaluation process, which includes a segmented classification task and an unsegmented recognition task.

Abstract: This document describes techniques for radio frequency (RF) based micro-motion tracking. These techniques enable even millimeter-scale hand motions to be tracked. To do so, radar signals are used from radar systems that, with conventional techniques, would only permit resolutions of a centimeter or more.

Abstract: Techniques and apparatuses are described that provide radar sensing for multiple applications. In an example aspect, middleware is coupled between multiple applications and a radar system. The middleware performs translation services, conflict resolution and/or resource management to configure the radar system in a manner that can provide radar sensing to at least a substantial subset of the applications during a given time interval. Also, the middleware can dynamically adjust the operation of the radar system as different applications request radar sensing. The middleware enables the radar system to provide radar sensing for a larger quantity of diverse applications without integrating additional radar systems into the computing device and/or without the applications needing to communicate and/or negotiate with each other. In this way, the middleware can expand the utilization of the radar system, thereby providing additional features to enhance the user experience.

Abstract: Techniques and apparatuses are described that facilitate ambient computing using a radar system. Compared to other smart devices that rely on a physical user interface, a smart device with a radar system can support ambient computing by providing an eye-free interaction and less cognitively demanding gesture-based user interface. The radar system can be designed to address a variety of challenges associated with ambient computing, including power consumption, environmental variations, background noise, size, and user privacy. The radar system uses an ambient-computing machine-learned module to quickly recognize gestures performed by a user up to at least two meters away. The ambient-computing machine-learned module is trained to filter background noise and have a sufficiently low false positive rate to enhance the user experience.

Abstract: This application is directed to an integrated multimodal neural network driven by a natural language prompt. A computer system obtains sensor data from a plurality of sensor devices disposed in a physical environment during a time duration. One or more information items are generated to characterize one or more signature events detected within the time duration in the sensor data. The computer system obtains a natural language prompt. In response to the natural language prompt, the computer system applies a large behavior model (e.g., a large language model, a data processing model) to process the one or more information items and the natural language prompt jointly and generate a multimodal output (e.g., textual statements, software code, an image or video, an information dashboard having a predefined format, a user interface, and a heatmap). The multimodal output associated with the sensor data is represented.

Abstract: This application is directed to integrated multimodal neural networks. A computer system obtains sensor data from a plurality of sensor devices during a time duration, and the plurality of sensor devices include at least two distinct senor types and are disposed in a physical environment. One or more signature events are detected in the sensor data, and one or more information items are generated to characterize the one or more signature events detected in the sensor data, independently of the sensor types of the sensor devices. A large behavior model is applied to process the one or more information items and generate a multimodal output associated with the sensor data. The multimodal output describes the signature events associated with the sensor data in one of a plurality of predefined output modalities. The multimodal output is presented according to the one of the plurality of predefined output modalities.

Abstract: This application is directed to compressing sensor data. A computer system obtains the sensor data from a plurality of sensor devices disposed in a physical environment during a time duration, and each sensor device corresponds to a temporal sequence of respective sensor samples. For each of the plurality of sensor devices, the temporal sequence of respective sensor samples is processed to generate an ordered sequence of respective sensor data features defining a respective parametric representation of the temporal sequence of respective sensor samples, independently of a sensor type of the respective sensor device. The computer system detects one or more signature events within the time duration based on the respective parametric representations of the plurality of sensor devices, and generates one or more information items characterizing the one or more signature events detected in the sensor data.

Abstract: Devices are provided that include radar circuits arranged to send and receive radar signals that can be used to, for example, detect gestures performed in the vicinity of the device. Arrangements of the circuits and associated antennas allow for the device to have no bezel or a minimal bezel.

Abstract: This document describes apparatuses and techniques for radar-enabled sensor fusion. In some aspects, a radar field is provided and reflection signals that correspond to a target in the radar field are received. The reflection signals are transformed to provide radar data, from which a radar feature indicating a physical characteristic of the target is extracted. Based on the radar features, a sensor is activated to provide supplemental sensor data associated with the physical characteristic. The radar feature is then augmented with the supplemental sensor data to enhance the radar feature, such as by increasing an accuracy or resolution of the radar feature. By so doing, performance of sensor-based applications, which rely on the enhanced radar features, can be improved.

Abstract: Techniques and apparatuses are described that facilitate ambient computing using a radar system. Compared to other smart devices that rely on a physical user interface, a smart device with a radar system can support ambient computing by providing an eye-free interaction and less cognitively demanding gesture-based user interface. The radar system can be designed to address a variety of challenges associated with ambient computing, including power consumption, environmental variations, background noise, size, and user privacy. The radar system uses an ambient-computing machine-learned module to quickly recognize gestures performed by a user up to at least two meters away. The ambient-computing machine-learned module is trained to filter background noise and have a sufficiently low false positive rate to enhance the user experience.

Abstract: Techniques and apparatuses are described that enable low-power radar. The described techniques enable a radar system to reduce overall power consumption, thereby facilitating incorporation and utilization of the radar system within power-limited devices. Power consumption is reduced through customization of the transmission or processing of radar signals within the radar system. During transmission, different duty cycles, transmit powers, or framing structures can be utilized to collect appropriate data based on detected activity in an external environment. During processing, different hardware or different radar pipelines can be utilized to appropriately analyze the radar data. Instead of disabling the radar system, the described techniques enable the radar system to continuously monitor a dynamic environment and maintain responsiveness while conserving power.

Abstract: Techniques and apparatuses are described that implement a smart-device-based radar system capable of detecting user gestures in the presence of saturation. In particular, a radar system employs machine learning to compensate for distortions resulting from saturation. This enables gesture recognition to be performed while the radar system's receiver is saturated. As such, the radar system can forgo integrating an automatic gain control circuit to prevent the receiver from becoming saturated. Furthermore, the radar system can operate with higher gains to increasing sensitivity without adding additional antennas. By using machine learning, the radar system's dynamic range increases, which enables the radar system to detect a variety of different types of gestures having small or large radar cross sections, and performed at various distances from the radar system.

Abstract: Techniques and apparatuses are described that implement electromagnetic vector sensors (EMVS) for a smart-device-based radar system. Instead of including an antenna array of similar antenna elements, the radar system includes two or more electromagnetic vector sensors. At least one of the electromagnetic vector sensors is used for transmission and at least another of the electromagnetic vector sensors is used for reception. Each electromagnetic vector sensor includes a group of antennas with different antenna patterns, orientations, and/or polarizations. An overall footprint of the two electromagnetic vector sensors (e.g., one for transmission and one for reception) can be smaller than antenna arrays used by other radar systems, thereby enabling the radar system to be implemented within space-constrained devices.

Abstract: This document describes techniques for radio frequency (RF) based micro-motion tracking. These techniques enable even millimeter-scale hand motions to be tracked. To do so, radar signals are used from radar systems that, with conventional techniques, would only permit resolutions of a centimeter or more.

Abstract: This document describes techniques and systems that enable a mobile device-based radar system for applying different power modes to a multi-mode interface. The techniques and systems include a user device having a radar system, and an interaction manager. The radar system generates a radar field, provides radar data, and operates at one of various different radar-power states. The user device analyzes the radar data to detect a presence or movement of a user within the radar field. Responsive to the detection, the radar system changes from a first radar-power state to a second radar-power state. Based on this change, the interaction manager selects a power mode, for a multi-mode interface, that corresponds to the second radar-power state, and applies the selected power mode to the multi-mode interface to provide a corresponding display via a display device.

Abstract: This document describes techniques for radio frequency (RF) based micro-motion tracking. These techniques enable even millimeter-scale hand motions to be tracked. To do so, radar signals are used from radar systems that, with conventional techniques, would only permit resolutions of a centimeter or more.

Abstract: Various embodiments wirelessly detect micro gestures using multiple antenna of a gesture sensor device. At times, the gesture sensor device transmits multiple outgoing radio frequency (RF) signals, each outgoing RF signal transmitted via a respective antenna of the gesture sensor device. The outgoing RF signals are configured to help capture information that can be used to identify micro-gestures performed by a hand. The gesture sensor device captures incoming RF signals generated by the outgoing RF signals reflecting off of the hand, and then analyzes the incoming RF signals to identify the micro-gesture.