Abstract: Disclosed are techniques for an artificial intelligence (AI) based recommendation system associated with treatment of a patient. The AI based recommendation system can be configured to receive three-dimensional (3D) image data that is associated with a patient, where the 3D image data includes volumetric data; and detect one or more anatomical treatment sites associated with the patient. The 3D image data may further be annotated with one or more of the detected anatomical treatment sites. By leveraging a foundation model with detailed context on flap harvesting, the AI based recommendation system may evaluate the annotated 3D image data, identify viable donor sites for flap harvesting, and provide recommendations for one or more viable donor sites for flap harvesting based on a treatment criteria for treatment of the patient. The recommendation may be evaluated and further refined by additional user based interaction with the AI based system.

Abstract: Disclosed are systems and methods that leverage machine learning techniques to provide information about items that are on display during a live event. A 3D volumetric capture system is used to construct a 3D representation of the event in real time. For example, depth and color data from multiple perspectives may be used to construct a 3D mesh of colored polygons that represent the scene. Users may view the live event with a 3D enabled virtual reality or augmented reality headset. Background subtraction may be used to isolate items of interest. In some configurations a machine learning model may process the 3D mesh to generate a textual description of items in the scene. The textual description may then be used to search for more information about a particular item. Additionally, or alternatively, the machine learning model may directly predict one or more items in the scene.

Abstract: 3D teleconferences use an array of cameras to generate a 3D model of a subject. During a calibration and registration process the pose of each camera may be adjusted. Similarly, camera settings such as focus depth and white balance may be modified. These changes are made to improve the quality of the 3D model generated from image data captured by the cameras. Many factors affect the quality of images captured by the cameras. For example, depth sensors may be affected by the skin tone of the subject. In some configurations, a machine learning model (ML model) is trained on adjustments to properties that affect 3D model quality. The resulting ML model may then be used to infer camera adjustments for a given set of subject attributes, camera properties, and/or environment properties.

Abstract: Disclosed are systems and methods that leverage machine learning techniques to provide information about items that are on display during a live event. A 3D volumetric capture system is used to construct a 3D representation of the event in real time. For example, depth and color data from multiple perspectives may be used to construct a 3D mesh of colored polygons that represent the scene. Users may view the live event with a 3D enabled virtual reality or augmented reality headset. Background subtraction may be used to isolate items of interest. In some configurations a machine learning model may process the 3D mesh to generate a textual description of items in the scene. The textual description may then be used to search for more information about a particular item. Additionally, or alternatively, the machine learning model may directly predict one or more items in the scene.

Abstract: In addition to an original prompt that is manually provided by a user, contextual information is sent to a generative AI to elicit a higher quality response. Sensors collect audio, video, physiological, cognitive, environmental, and digital data from the user. Machine-learning models evaluate the sensor data to infer the emotional state of the user. The emotional state is used to augment the original prompt with contextual information. The augmented prompt is fed into the generative AI to make it context-aware. Accordingly, the generative AI can automatically pick up on non-verbal cues that the user did not manually articulate in the original prompt. Just as a human-to-human conversation involves a combination of verbal and non-verbal communications, the present concepts enable the generative AI to also leverage non-verbal communication when interacting with human users.

Abstract: An example magnetic actuator includes first and second flexible magnetic layers and a control circuit. The first flexible magnetic layer is configured to support a first magnetic dipole. Arranged in slidable contact with the first flexible magnetic layer, the second flexible magnetic layer is configured to support a second magnetic dipole. The control circuit is configured to controllably form at least the first magnetic dipole and thereby modify a force of interaction between the first and second flexible magnetic layers.

Abstract: An example magnetic actuator includes first and second flexible magnetic layers and a control circuit. The first flexible magnetic layer is configured to support a first magnetic dipole. Arranged in slidable contact with the first flexible magnetic layer, the second flexible magnetic layer is configured to support a second magnetic dipole. The control circuit is configured to controllably form at least the first magnetic dipole and thereby modify a force of interaction between the first and second flexible magnetic layers.