Abstract: Systems, apparatuses and methods may provide for technology that processes an inference workload in a first subset of layers of a neural network that prevents or inhibits data dependent branch operations, conducts an exit determination as to whether an output of the first subset of layers satisfies one or more exit criteria, and selectively bypasses processing of the output in a second subset of layers of the neural network based on the exit determination. The technology may also speculatively initiate the processing of the output in the second subset of layers while the exit determination is pending. Additionally, when the inference workloads include a plurality of batches, the technology may mask one or more of the plurality of batches from processing in the second subset of layers.

Abstract: Systems, apparatuses and methods may provide for technology that processes an inference workload in a first subset of layers of a neural network that prevents or inhibits data dependent branch operations, conducts an exit determination as to whether an output of the first subset of layers satisfies one or more exit criteria, and selectively bypasses processing of the output in a second subset of layers of the neural network based on the exit determination. The technology may also speculatively initiate the processing of the output in the second subset of layers while the exit determination is pending. Additionally, when the inference workloads include a plurality of batches, the technology may mask one or more of the plurality of batches from processing in the second subset of layers.

Abstract: Systems, apparatuses and methods may provide for technology that processes an inference workload in a first subset of layers of a neural network that prevents or inhibits data dependent branch operations, conducts an exit determination as to whether an output of the first subset of layers satisfies one or more exit criteria, and selectively bypasses processing of the output in a second subset of layers of the neural network based on the exit determination. The technology may also speculatively initiate the processing of the output in the second subset of layers while the exit determination is pending. Additionally, when the inference workloads include a plurality of batches, the technology may mask one or more of the plurality of batches from processing in the second subset of layers.

Abstract: Systems, apparatuses and methods may provide for technology that processes an inference workload in a first subset of layers of a neural network that prevents or inhibits data dependent branch operations, conducts an exit determination as to whether an output of the first subset of layers satisfies one or more exit criteria, and selectively bypasses processing of the output in a second subset of layers of the neural network based on the exit determination. The technology may also speculatively initiate the processing of the output in the second subset of layers while the exit determination is pending. Additionally, when the inference workloads include a plurality of batches, the technology may mask one or more of the plurality of batches from processing in the second subset of layers.

Abstract: An augmented reality (AR) device includes a 3D video camera to capture video images and corresponding depth information, a display device to display the video data, and an AR module to add a virtual 3D model to the displayed video data. A depth mapping module generates a 3D map based on the depth information, a dynamic scene recognition and tracking module processes the video images and the 3D map to detect and track a target object within a field of view of the 3D video camera, and an augmented video rendering module renders an augmented video of the virtual 3D model dynamically interacting with the target object. The augmented video is displayed on the display device in real time. The AR device may further include a context module to select the virtual 3D model based on context data comprising a current location of the augmented reality device.

Abstract: A mechanism is described for facilitating smart measurement of body dimensions despite loose clothing and/or other obscurities according to one embodiment. A method of embodiments, as described herein, includes capturing, by one or more capturing/sensing components of a computing device, a scan of a body of a user, and computing one or more primary measurements relating to one or more primary areas of the body, where the one or more primary measurements are computed based on depth data of the one or more primary areas of the body, where the depth data is obtained from the scan. The method may further include receiving at least one of secondary measurements and a three-dimensional (3D) avatar of the body based on the primary measurements, and preparing a report including body dimensions of the body based on at least one of the secondary measurements and the 3D avatar, and presenting the report at a display device.

Abstract: A mechanism is described for facilitating smart measurement of body dimensions despite loose clothing and/or other obscurities according to one embodiment. A method of embodiments, as described herein, includes capturing, by one or more capturing/sensing components of a computing device, a scan of a body of a user, and computing one or more primary measurements relating to one or more primary areas of the body, where the one or more primary measurements are computed based on depth data of the one or more primary areas of the body, where the depth data is obtained from the scan. The method may further include receiving at least one of secondary measurements and a three-dimensional (3D) avatar of the body based on the primary measurements, and preparing a report including body dimensions of the body based on at least one of the secondary measurements and the 3D avatar, and presenting the report at a display device.

Abstract: Various systems and methods for providing a wellness mirror are provided herein. A system for providing a wellness mirror includes a display; a modeler to receive depth images from a depth camera that is communicatively coupled to the system, and provide a model of a subject in the depth images; a health profiler to analyze the model and produce a health and wellness analysis; and a user interface to present the health and wellness analysis on the display.

Abstract: A mechanism is described for facilitating smart measurement of body dimensions despite loose clothing and/or other obscurities according to one embodiment. A method of embodiments, as described herein, includes capturing, by one or more capturing/sensing components of a computing device, a scan of a body of a user, and computing one or more primary measurements relating to one or more primary areas of the body, where the one or more primary measurements are computed based on depth data of the one or more primary areas of the body, where the depth data is obtained from the scan. The method may further include receiving at least one of secondary measurements and a three-dimensional (3D) avatar of the body based on the primary measurements, and preparing a report including body dimensions of the body based on at least one of the secondary measurements and the 3D avatar, and presenting the report at a display device.

Abstract: An augmented reality (AR) device includes a 3D video camera to capture video images and corresponding depth information, a display device to display the video data, and an AR module to add a virtual 3D model to the displayed video data. A depth mapping module generates a 3D map based on the depth information, a dynamic scene recognition and tracking module processes the video images and the 3D map to detect and track a target object within a field of view of the 3D video camera, and an augmented video rendering module renders an augmented video of the virtual 3D model dynamically interacting with the target object. The augmented video is displayed on the display device in real time. The AR device may further include a context module to select the virtual 3D model based on context data comprising a current location of the augmented reality device.

Abstract: An augmented reality (AR) device includes a 3D video camera to capture video images and corresponding depth information, a display device to display the video data, and an AR module to add a virtual 3D model to the displayed video data. A depth mapping module generates a 3D map based on the depth information, a dynamic scene recognition and tracking module processes the video images and the 3D map to detect and track a target object within a field of view of the 3D video camera, and an augmented video rendering module renders an augmented video of the virtual 3D model dynamically interacting with the target object. The augmented video is displayed on the display device in real time. The AR device may further include a context module to select the virtual 3D model based on context data comprising a current location of the augmented reality device.

Abstract: Examples are disclosed for determining an interestingness score for one or more areas of interest included in a display element such as a static image or a motion video. In some examples, an interestingness score may be determined based on eye tracking or gaze information gathered while an observer views the display element.

Abstract: Apparatuses, systems, media and/or methods may involve creating content. A property component may be added to a media object to impart one or more of a perceptual property or a contextual property to the media object. The property component may be added responsive to an operation by a user that is independent of a direct access by the user to computer source code. An event corresponding to the property component may be mapped with an action for the media object. The event may be mapped with the action responsive to an operation by a user that is independent of a direct access by the user to computer source code. A graphical user interface may be rendered to create the content. In addition, the media object may be modified based on the action in response to the event when content created including the media object is utilized.

Abstract: Systems, apparatuses and methods may provide for technology that processes an inference workload in a first subset of layers of a neural network that prevents or inhibits data dependent branch operations, conducts an exit determination as to whether an output of the first subset of layers satisfies one or more exit criteria, and selectively bypasses processing of the output in a second subset of layers of the neural network based on the exit determination. The technology may also speculatively initiate the processing of the output in the second subset of layers while the exit determination is pending. Additionally, when the inference workloads include a plurality of batches, the technology may mask one or more of the plurality of batches from processing in the second subset of layers.

Abstract: Computer-readable storage media, computing devices and methods are discussed herein. In embodiments, a computing device may include one or more display devices, a digital content module coupled with the one or more display devices, and an augmentation module coupled with the digital content module and the one or more display devices. The digital content module may be configured to cause a portion of textual content to be rendered on the one or more display devices. The textual content may be associated with a digital scene that may be utilized to augment the textual content. The augmentation module may be configured to dynamically adapt the digital scene, based at least in part on a real-time video feed, to be rendered on the one or more display devices to augment the textual content. Other embodiments may be described and/or claimed.

Abstract: Various systems and methods for generating and outputting dynamic depth-based content in a virtual reality (VR) environment are described. For example, a technique for generating location-customized content in VR may be implemented by electronic operations that: detect an object from image data of a real-world environment; identify the real-world location of the object relative to a viewing position with a VR device; identify a corresponding virtual location for a selected virtual object; and display the virtual object at the corresponding virtual location in VR. The image data may be generated from an image sensor and a depth sensor that captures three-dimensional aspects of the real-world environment. Based on the type and characteristics of the real-world object, a corresponding virtual object may be presented and interacted with, to allow a human user to avoid real-world obstacles or other objects during a VR session.

Abstract: Various embodiments are generally directed to techniques for providing an augmented reality view in which eye movements are employed to identify items of possible interest for which indicators are visually presented in the augmented reality view. An apparatus to present an augmented reality view includes a processor component; a presentation component for execution by the processor component to visually present images captured by a camera on a display, and to visually present an indicator identifying an item of possible interest in the captured images on the display overlying the visual presentation of the captured images; and a correlation component for execution by the processor component to track eye movement to determine a portion of the display gazed at by an eye, and to correlate the portion of the display to the item of possible interest. Other embodiments are described and claimed.

Abstract: Techniques related to 3D image capture with dynamic cameras.

Abstract: Systems, apparatuses, and/or methods to augment reality. An object identifier may identify an object in a field of view of a user that includes a reflection of the user from a reflective surface, such as a surface of a traditional mirror. In addition, a reality augmenter may generate an augmented reality object based on the identification of the object. In one example, eyeglasses including a relatively transparent display screen may be coupled with an image capture device on the user and the augmented reality object may be observable by the user on the transparent display screen when the user wears the eyeglasses. A localizer may position the augmented reality object on the transparent display screen relative to the reflection of the user that passes though the transparent display screen during natural visual perception of the reflection by the user.

Abstract: An augmented reality (AR) device includes a 3D video camera to capture video images and corresponding depth information, a display device to display the video data, and an AR module to add a virtual 3D model to the displayed video data. A depth mapping module generates a 3D map based on the depth information, a dynamic scene recognition and tracking module processes the video images and the 3D map to detect and track a target object within a field of view of the 3D video camera, and an augmented video rendering module renders an augmented video of the virtual 3D model dynamically interacting with the target object. The augmented video is displayed on the display device in real time. The AR device may further include a context module to select the virtual 3D model based on context data comprising a current location of the augmented reality device.