Abstract: Systems, apparatus, articles of manufacture, and methods to detect anomalies in three-dimensional (3D) images are disclosed. Example apparatus disclosed herein generate a first two-dimensional (2D) anomaly map corresponding to a first 2D image slice of a 3D image, the first 2D image slice corresponding to a first axis of the 3D image. Disclosed example apparatus also generate a second 2D anomaly map corresponding to a second 2D image slice of the 3D image, the second 2D image slice corresponding to a second axis of the 3D image. Disclosed example apparatus further generate a 3D anomaly volume based on the first 2D anomaly map and the second 2D anomaly detection, the 3D anomaly volume corresponding to the 3D image.

Abstract: Features extracted from one or more layers of a trained deep neural network (DNN) are used to detect out-of-distribution (OOD) data, such as anomalies. An OOD detection process includes inputting a feature output from a layer of the DNN into a trained autoencoder that applies an encoding function followed by a decoding function, resulting in a reconstructed feature. A feature reconstruction error is calculated based on a difference between the reconstructed feature and the original feature output from the DNN. The OOD detection process may further include calculating a score based on the feature reconstruction error and generating a visual representation of the feature reconstruction error.

Abstract: An example apparatus includes interface circuitry, machine-readable instructions, and at least one processor circuit to be programmed by the machine-readable instructions to extract neural network model features from deployment data, identify out-of-distribution data based on the neural network model features, identify samples with the out-of-distribution data to generate one or more scores associated with post-deployment data drift, and classify post-deployment data based on the one or more scores.

Abstract: A method for identifying a tool anomaly of an printed circuit board (PCB) manufacturing process comprising a plurality of phases, the method comprising the steps of: obtaining image data of at least one tool of the PCB manufacturing process; inputting the image data to a machine learning module, the machine learning module configured to perform the following steps: extracting, from the image data, a tool feature image data of the at least one tool; classifying the image data into a phase of the plurality of phases; and determining, based on the classified image data and the tool feature image data, an anomaly state of the at least one tool.

Abstract: Techniques are disclosed for using neural network architectures to estimate predictive uncertainty measures, which quantify how much trust should be placed in the deep neural network (DNN) results. The techniques include measuring reliable uncertainty scores for a neural network, which are widely used in perception and decision-making tasks in automated driving. The uncertainty measurements are made with respect to both model uncertainty and data uncertainty, and may implement Bayesian neural networks or other types of neural networks.

Abstract: Methods and apparatus are disclosed herein for computation and compression efficiency in distributed video analytics. Example apparatus disclosed herein are to identify a key frame and a non-key frame in a video frame sequence input to a neural network at a client server, determine motion information between the key frame and the non-key frame based on optical flow, and determine a frame feature representation based on the motion information reconstructed at an edge server, the motion information including feature warping residual errors.

Abstract: Features extracted from one or more layers of a trained deep neural network (DNN) are used to detect out-of-distribution (OOD) data, such as anomalies. An OOD detection process includes transforming a feature output from a layer of the DNN from a relatively high-dimensional feature space to a lower-dimensional space, and then performing a reverse transformation back to the higher-dimensional feature space, resulting in a reconstructed feature. A feature reconstruction error is calculated based on a difference between the reconstructed feature and the original feature output from the DNN. The OOD detection process may further include calculating a score based on the feature reconstruction error and generating a visual representation of the feature reconstruction error.

Abstract: Disclosed is an example solution to perform source-free active adaptation to distributional shifts for machine learning. The example solution includes: interface circuitry; programmable circuitry; and instructions to cause the programmable circuitry to: perform a first training of a neural network on a baseline data set associated with a first data distribution; compare data of a shifted data set to a threshold uncertainty value, wherein the threshold uncertainty value is associated with a distributional shift between the baseline data set and the shifted data set; generate a shifted data subset including items of the shifted dataset that satisfy the threshold uncertainty value; and perform a second training of the neural network based on the shifted data subset.

Abstract: Vehicle navigation control systems in autonomous driving rely on accurate predictions of objects within the vicinity of the vehicle to appropriately control the vehicle safely through its surrounding environment. Accordingly this disclosure provides methods and devices which implement mechanisms for obtaining contextual variables of the vehicle's environment for use in determining the accuracy of predictions of objects within the vehicle's environment.

Abstract: Disclosures herein may be directed to a method, technique, or apparatus directed to a computer-assisted or autonomous driving (CA/AD) vehicle that includes a system controller, disposed in a first CA/AD vehicle, to manage a collaborative three-dimensional (3-D) map of an environment around the first CA/AD vehicle, wherein the system controller is to receive, from another CA/AD vehicle proximate to the first CA/AD vehicle, an indication of at least a portion of another 3-D map of another environment around both the first CA/AD vehicle and the another CA/AD vehicle and incorporate the at least the portion of the 3-D map proximate to the first CA/AD vehicle and the another CA/AD vehicle into the 3-D map of the environment of the first CA/AD vehicle managed by the system controller.

Abstract: Various systems and methods for providing variable bitrate compression for split deep neural network (DNN) computing are described herein. A system may be configured to manage a split DNN, the split DNN configured to operate on a compute system and a second system over a communication network. The system may access a performance metric; determine, based on the performance metric, a split point of the split DNN, the split point defining a head portion of the split DNN and a tail portion of the split DNN; determine, based on the performance metric, a bottleneck layer configuration for a bottleneck layer at the split point, the bottleneck layer including a bottleneck encoder and a bottleneck decoder; execute the head portion of the DNN and the bottleneck encoder on the compute system; and recurrently access an updated performance metric and determine a revised split point or a revised bottleneck layer configuration based on the updated performance metric.

Abstract: A safety system for a vehicle may include one or more processors configured to determine uncertainty data indicating uncertainty in one or more predictions from a driving model during operation of a vehicle; change or update one or more of the driving model parameters to one or more changed or updated driving model parameters based on the determined uncertainty data; and provide the one or more changed or updated driving model parameters to a control system of the vehicle for controlling the vehicle to operate in accordance with the driving model including the one or more changed or updated driving model parameters.

Abstract: An apparatus to facilitate ensemble learning for deep feature defect detection is disclosed. The apparatus includes one or more processors to receive a deep feature vector from a feature extractor of an ensemble learning system, the deep feature vector extracted from input data; cluster the deep feature vector into a plurality of clusters based on a distance into the plurality of clusters; execute a probabilistic machine learning model corresponding to a cluster of the plurality of clusters to which the deep feature vector is clustered; and detect whether the deep feature vector comprises a defect based on an output of execution of the probabilistic machine learning model.

Abstract: Embodiments include apparatuses, systems, and methods for a computer-aided or autonomous driving (CA/AD) system to detect an anomalous image associated with image data from one or more cameras of a computer-aided or autonomous driving (CA/AD) vehicle. Embodiments may include a sensor interface disposed in the CA/AD vehicle to receive, from the one or more cameras, a stream of image data including single view image data captured by the one or more cameras or multi-view image data collaboratively captured by multiple ones of the one or more cameras. In embodiments, a consistency analysis unit disposed in the CA/AD vehicle is coupled to the sensor interface to perform a consistency check on pixel-level data using single view or multi-view geometric methods to determine whether the image data includes an anomalous image. Other embodiments may also be described and claimed.

Abstract: A human-robot collaboration system, including at least one processor; and a non-transitory computer-readable storage medium including instructions that, when executed by the at least one processor, cause the at least one processor to: predict a human atomic action based on a probability density function of possible human atomic actions for performing a predefined task; and plan a motion of the robot based on the predicted human atomic action.

Abstract: Estimating a head pose may include obtaining sensor data corresponding to a head and at least a portion of the body of a human subject and determining an estimate of a three-dimensional (3D) body pose using the obtained sensor data. The estimation can further include generating a first rendering of at least the human subject's head using the obtained sensor data and generating a plurality of head pose sample data sets by applying the estimated 3D body pose to a head-pose generative model. Further, the head pose estimation can include generating a plurality of second renderings respectively from each of the plurality of head pose sample data sets; determining which of the plurality of second renderings is closest to the first rendering; and selecting the second rendering determined to be closest to the first rendering.

Abstract: Methods, apparatus, systems and articles of manufacture are disclosed to facilitate continuous learning. An example apparatus includes a trainer to train a first Bayesian neural network (BNN) and a second BNN, the first BNN associated with a first weight distribution and the second BNN associated with a second weight distribution. The example apparatus includes a weight determiner to determine a first sampling weight associated with the first BNN and a second sampling weight associated with the second BNN. The example apparatus includes a network sampler to sample at least one of the first weight distribution or the second weight distribution based on a pseudo-random number, the first sampling weight, and the second sampling weight. The example apparatus includes an inference controller to generate an ensemble weight distribution based on the sample.

Abstract: Methods, systems, and apparatus to obtain well-calibrated uncertainty in probabilistic deep neural networks are disclosed. An example apparatus includes a loss function determiner to determine a differentiable accuracy versus uncertainty loss function for a machine learning model, a training controller to train the machine learning model, the training including performing an uncertainty calibration of the machine learning model using the loss function, and a post-hoc calibrator to optimize the loss function using temperature scaling to improve the uncertainty calibration of the trained machine learning model under distributional shift.

Abstract: Methods, apparatus, systems and articles of manufacture are disclosed to facilitate knowledge sharing among neural networks. An example apparatus includes a trainer to train, at a first computing system, a first Bayesian Neural Network (BNN) on a first subset of training data to generate a first weight distribution, and train, at a second computing system, a second BNN on a second subset of the training data to generate a second weight distribution, the second subset of the training data different from the first subset of training data. The example apparatus includes a knowledge sharing controller to generate a third BNN based on the first weight distribution and the second weight distribution.

Abstract: Techniques are disclosed for using neural network architectures to estimate predictive uncertainty measures, which quantify how much trust should be placed in the deep neural network (DNN) results. The techniques include measuring reliable uncertainty scores for a neural network, which are widely used in perception and decision-making tasks in automated driving. The uncertainty measurements are made with respect to both model uncertainty and data uncertainty, and may implement Bayesian neural networks or other types of neural networks.