Abstract: Methods, systems, and computer programs are presented for implementing Personalized Mobility as a Service (PMaaS) to improve transportation services delivery. One storage medium includes instructions for detecting, by a mobility as a service (MaaS) system, a request for a trip from a user device of a user. The storage medium further includes instructions for mapping, using a model executing on the machine, the user to a persona from a plurality of persona models. Each persona model has one or more characteristics associated with users of the MaaS system. Further yet, the storage medium includes instructions for determining trip parameters for the trip based on the persona mapped to the user, the trip parameters defining one or more trip segments for the trip, and instructions for providing trip parameters to the user device.

Abstract: In one embodiment, an apparatus comprises a memory and a processor. The memory is to store sensor data, wherein the sensor data is captured by a plurality of sensors within an educational environment. The processor is to: access the sensor data captured by the plurality of sensors: identify a student within the educational environment based on the sensor data: detect a plurality of events associated with the student based on the sensor data, wherein each event is indicative of an attention level of the student within the educational environment: generate a report based on the plurality of events associated with the student; and send the report to a third party associated with the student.

Abstract: An apparatus may include a memory to store a recorded video. The apparatus may further include an interface to receive at least one set of sensor information based on sensor data that is recorded concurrently with the recorded video and a video clip creation module to identify a sensor event from the at least one set of sensor information and to generate a video clip based upon the sensor event, the video clip comprising video content from the recorded video that is synchronized to the sensor event.

Abstract: Methods and apparatus to operate closed-lid portable computers are disclosed. An example portable computer includes a first display on a lid of the portable computer, the first display to be deactivated when the lid is in a closed position; a second display distinct from the first display, the second display to be visible when the lid is in the closed position; instructions; and processor circuitry to execute the instructions to cause activation of the first display in response to a user interaction with the second display while the lid is in the closed position.

Abstract: Machine learning models can process a video and generate outputs such as action segmentation assigning portions of the video to a particular action, or action classification assigning an action class for each frame of the video. Some machine learning models can accurately make predictions for short videos but may not be particularly suited for performing action segmentation for long duration, structured videos. An effective machine learning model may include a hybrid architecture involving a temporal convolutional network and a bi-directional graph neural network. The machine learning model can process long duration structured videos by using a temporal convolutional network as a first pass action segmentation model to generate rich, frame-wise features. The frame-wise features can be converted into a graph having forward edges and backward edges. A graph neural network can process the graph to refine a final fine-grain per-frame action prediction.

Abstract: Methods and apparatus to operate closed-lid portable computers are disclosed. An example portable computer includes a first display on a lid of the portable computer, the first display to be deactivated when the lid is in a closed position; a second display distinct from the first display, the second display to be visible when the lid is in the closed position; instructions; and processor circuitry to execute the instructions to cause activation of the first display in response to a user interaction with the second display while the lid is in the closed position.

Abstract: Systems and methods for multi-modal user device authentication are disclosed. An example electronic device includes a first sensor, a microphone, a first camera, and a confidence analyzer to authenticate a subject as the authorized user in response to a user presence detection analyzer detecting a presence of the subject and one or more of (a) an audio data analyzer detecting a voice of an authorized user or (b) an image data analyzer detecting a feature of the authorized user. The example electronic device includes a processor to cause the electronic device to move from a first power state to a second power state in response to the confidence analyzer authenticating the user as the authorized user. The electronic device is to consume a greater amount of power in the second power state than the first power state.

Abstract: Systems and methods for multi-modal user device authentication are disclosed. An example electronic device includes a first sensor, a microphone, a first camera, and a confidence analyzer to authenticate a subject as the authorized user in response to a user presence detection analyzer detecting a presence of the subject and one or more of (a) an audio data analyzer detecting a voice of an authorized user or (b) an image data analyzer detecting a feature of the authorized user. The example electronic device includes a processor to cause the electronic device to move from a first power state to a second power state in response to the confidence analyzer authenticating the user as the authorized user. The electronic device is to consume a greater amount of power in the second power state than the first power state.

Abstract: System and techniques for tracking caloric expenditure using sensor driven fingerprints are described herein. A set of outputs may be obtained from a plurality of sensors. A fingerprint may be generated using the set of outputs. The fingerprint may correspond to an activity observed by the plurality of sensors. The generated fingerprint may be compared to a set of fingerprints stored in a database. Each fingerprint of the set of fingerprints may correspond to a respective caloric expenditure. A caloric expenditure may be calculated for the activity based on the comparison. An exercise profile of a user may be updated using the caloric expenditure.

Abstract: Methods and apparatus to operate closed-lid portable computers are disclosed. An example portable compute device includes: a microphone; a speaker; a first camera to face a first direction; and a second camera to face a second direction, the second direction opposite the first direction. The compute device further includes communications circuitry; a first display; a second display separate from the first display; and a hinge to enable the first display to rotate relative to the second display between an open position and a closed position. At least a portion of the second display is capable of being visible when the first display is rotated about the hinge to the closed position. The portion of the second display is multiple times longer in a third direction than in a fourth direction perpendicular to the third direction, the third direction extending parallel to an axis of rotation of the hinge.

Abstract: Methods and apparatus to operate closed-lid portable computers are disclosed. An example portable compute device includes: a microphone; a speaker; a first camera to face a first direction; and a second camera to face a second direction, the second direction opposite the first direction. The compute device further includes communications circuitry; a first display; a second display separate from the first display; and a hinge to enable the first display to rotate relative to the second display between an open position and a closed position. At least a portion of the second display is capable of being visible when the first display is rotated about the hinge to the closed position. The portion of the second display is multiple times longer in a third direction than in a fourth direction perpendicular to the third direction, the third direction extending parallel to an axis of rotation of the hinge.

Abstract: Methods and apparatus to operate closed-lid portable computers are disclosed. An example portable computer includes a first display on a lid of the portable computer, the first display to be deactivated when the lid is in a closed position; a second display distinct from the first display, the second display to be visible when the lid is in the closed position; instructions; and processor circuitry to execute the instructions to cause activation of the first display in response to a user interaction with the second display while the lid is in the closed position.

Abstract: Logic may determine a specific performance of a neural network based on an event and may present the specific performance to provide a user with an explanation of the inference by a machine learning model such as a neural network. Logic may determine a first activation profile associated with the event, the first activation profile based on activation of nodes in one or more layers of the neural network during inference to generate an output. Logic may correlate the first activation profile against a second activation profile associated with a first training sample of training data. Logic may determine that the first training sample is associated with the event based on the correlation. Logic may output an indicator to identify the first training sample as being associated with the event.

Abstract: Disclosed is a technical solution to process a video that captures actions to be performed for completing a task based on a chronological sequence of stages within the task. An example system may identify an action sequence from an instruction for the task. The system inputs the action sequence into a trained model (e.g., a recurrent neural network), which outputs the chronological sequence of stages. The RNN may be trained through self-supervised learning. The system may input the video and the chronological sequence of stages into another trained model, e.g., a temporal convolutional network. The other trained model may include hidden layers arranged before an attention layer. The hidden layers may extract features from the video and feed the features into the attention layer. The attention layer may determine attention weights of the features based on the chronological sequence of stages.

Abstract: The systems and methods disclosed herein provide determination of an orientation of a feature towards a reference target. As a non-limiting example, a system consistent with the present disclosure may include a processor, a memory, and a single camera affixed to the ceiling of a room occupied by a person. The system may analyze images from the camera to identify any objects in the room and their locations. Once the system has identified an object and its location, the system may prompt the person to look directly at the object. The camera may then record an image of the user looking at the object. The processor may analyze the image to determine the location of the user's head and, combined with the known location of the object and the known location of the camera, determine the direction that the user is facing. This direction may be treated as a reference value, or “ground truth.” The captured image may be associated with the direction, and the combination may be used as training input into an application.

Abstract: Disclosed is a technical solution to calibrate confidence scores of classification networks. A classification network has been trained to receive an input and output a label of the input that indicates a class of the input. The classification network also outputs a confidence score of the label, which indicates a likelihood of the input falling into the class, i.e., a confidence level of the classification network that the label is correct. To calibrate the confidence of the classification network, a logit transformation function may be added into the classification network. The logic transformation function may be an entropy-based function and have learnable parameters, which may be trained by inputting calibration samples into the classification network and optimizing a negative log likelihood based on the labels generated by the classification network and ground-truth labels of the calibration samples. The trained logic transformation function can be used to compute reliable confidence scores.

Abstract: Systems, apparatuses, and methods include technology that generates final frame predictions for a first plurality of frames of a video, where the first plurality of frames is associated with unlabeled data. The technology predicts an ordered list of actions for the first plurality of frames based on the final frame predictions, and temporally aligning the ordered list of actions to the final frame predictions to generate labels.

Abstract: Technologies for time-delayed augmented reality (AR) presentations includes determining a location of a plurality of user AR systems located within the presentation site and determining a time delay of an AR sensory stimulus event of an AR presentation to be presented in the presentation site for each user AR system based on the location of the corresponding user AR system within the presentation site. The AR sensory stimulus event is presented to each user AR system based on the determined time delay associated with the corresponding user AR system. Each user AR system generates the AR sensory stimulus event based on a timing parameter that defines the time delay for the corresponding user AR system such that the generation of the AR sensory stimulus event is time-delayed based on the location of the user AR system within the presentation site.

Abstract: Systems, apparatuses, and methods include technology that identifies, with a neural network, that a predetermined amount of a first action is completed at a first portion of a plurality of portions. A subset of the plurality of portions collectively represents the first action. The technology generates a first loss based on the predetermined amount of the first action being identified as being completed at the first portion. The technology updates the neural network based on the first loss.

Abstract: Examples disclosed herein provide real-time language learning within a smart space. An example system includes a sensor; object detection software to identify a first object and a second object in an environment based on an output of the sensor; assign a first weight to the first object and a second weight to the second object; perform a comparison of the first weight and the second weight; and select the first object to be associated with a second language output based on the comparison; context determination software to determine a second language context based on the output of the sensor; linguistic analysis software to associate the first object with a second language based on the second language context; and prompt generation software to cause the second language output for the first object in the second language to be presented.