Abstract: A method for training a model for face recognition is provided. The method forward trains a training batch of samples to form a face recognition model w(t), and calculates sample weights for the batch. The method obtains a training batch gradient with respect to model weights thereof and updates, using the gradient, the model w(t) to a face recognition model what(t). The method forwards a validation batch of samples to the face recognition model what(t). The method obtains a validation batch gradient, and updates, using the validation batch gradient and what(t), a sample-level importance weight of samples in the training batch to obtain an updated sample-level importance weight. The method obtains a training batch upgraded gradient based on the updated sample-level importance weight of the training batch samples, and updates, using the upgraded gradient, the model w(t) to a trained model w(t+1) corresponding to a next iteration.

Abstract: A system for analyzing medical images to detect and classify a medical condition, the system includes an input for receiving a medical image, a convolutional neural network coupled to the input and configured to analyze the medical image to generate a prediction including a probability of the presence of the medical condition in the medical image and an atlas creation module coupled to the convolutional neural network and configured to generate an atlas comprising a set of image features and a set of training images. Each image feature is assigned with at least one training image associated with the medical condition. The system further includes a prediction basis selection module coupled to the convolutional neural network and the atlas and configured to create a prediction basis for the prediction generated by the convolutional neural network.

Abstract: Computer-implemented methods, systems, apparatus, and computer-readable media (transitory or non-transitory) are described herein for detecting eye tracking calibration errors. In various examples, respective poses of first and second eyes of a wearer of a head-mounted display (“HMD”) may be determined. A portion of a first image rendered on a first display of the HMD may be selected based on the pose of the first eye of the wearer. The selected portion of the first image may be matched to a portion of a second image rendered on a second display of the HMD. An eye tracking calibration error may be determined based on the pose of the second eye of the wearer and a location of the portion of the second image.

Abstract: A vehicular trailer angle detection system includes a camera disposed at a rear portion of a vehicle. The system determines, via processing of frames of image data captured by the camera, features of a trailer present rearward of the vehicle and hitched to the vehicle by determining features that have similar position changes between a current frame of image data captured by the camera and a previous frame of image data captured by the camera. Responsive to movement of the trailer relative to the vehicle, and via processing of captured frames of image data, the system tracks determined features over multiple captured frames of image data for different positions of the trailer relative to the vehicle. The system determines angle of the trailer relative to the vehicle based at least in part on tracking of determined features of the trailer over multiple captured frames of image data.

Abstract: A system includes a reconstructor (314) configured to reconstruct cone beam projection data to generate cone beam artifact corrected short scan cone beam volumetric image data. A method includes reconstructing, with a reconstructor, cone beam projection data to generate cone beam artifact corrected short scan cone beam volumetric image data. A computer-readable storage medium storing computer executable instructions which when executed by a processor of a computer cause the processor to: reconstruct cone beam projection data to generate cone beam artifact corrected short scan cone beam volumetric image data.

Abstract: A method for creating generated formatted data, that includes receiving, by a sequential data generator, raw data, where the raw data includes formation data at a drilling environment, processing the raw data to obtain generated recommendation data, where the generated recommendation data includes a proposed drilling location, and creating the generated formatted data, where the generated formatted data includes the generated recommendation data.

Abstract: Systems and methods for sender profile and/or recipient profile disambiguation and/or confirmation are disclosed. In instances where a sender profile is not indicated by a user sending a communication from a communal device, heuristic data may be utilized to infer the sender profile. Similar heuristic data may also be used when selection of the sender profile is associated with a low confidence level. Heuristic data may also be used to infer the recipient profile when the user does not indicate the recipient profile or when selection of the recipient profile is associated with a low confidence. Various confirmations may result from the sender and recipient profile disambiguation.

Abstract: An Artificial Intelligence (AI) based document processing and validation system identifies anomalies such as errors, fraud, and duplicates of received documents and enables automatic actions for valid documents using machine learning (ML) techniques. The received documents are processed for determining probabilities for errors, fraud, and duplicates. A validation worklist is generated with the documents arranged in descending order of the probabilities and invalid documents with higher probabilities are flagged for review while the valid documents with lower probabilities are further processed for the execution of automatic actions. The feedback from the invalid document review is used to further train the models in determining the probabilities.

Abstract: An image resynthesis system, a system for training a gap filling module to be used in the image resynthesis system, an image resynthesis method, a computer program product, and a computer-readable medium are provided. The image resynthesis system comprises a source image input module, a forward warping module predicting, for each source image pixel of a source image, a corresponding position in a target image, and predicting a forward warping field which is aligned with the source image, and a gap filling module filling in gaps resulting from application of the forward warping module.

Abstract: The present disclosure is directed to an apparatus and method for data analysis for use in data classification via training of a recurrent neural network to identify features from limited reference sets. Based on a one-shot learning algorithm, the method includes selecting a subset of reference data and training a classifier with the selected data. This small subset of reference data can be iteratively tuned to enhance classification of the data according to the desired output of the method. The apparatus may be configured to allow a user to interactively select a subset of reference data which is used to train the classifier and to evaluate classifier performance.

Abstract: An image processing apparatus in an embodiment includes a FIFO memory, a plurality of line buffers, an image processing circuit, and a control circuit. The plurality of line buffers store data inputted from a plurality of cameras. The image processing circuit performs predetermined image processing on the data stored in the plurality of line buffers. The control circuit performs control, according to an output control signal, such that output of data to the plurality of line buffers is stopped and the data stopped from being outputted is stored in the FIFO memory.

Abstract: An example method of decoding a point cloud includes selecting, from a plurality of predefined prediction modes, a prediction mode for performing predictive geometry coding of a position of a current node of the point cloud, wherein the plurality of prediction modes includes at least: a zero prediction mode, and a delta prediction mode; responsive to selecting the zero prediction mode: determining a radius, an azimuth, and a laser index of a parent node of the current node; inferring an azimuth and a laser index of a predicted position of the current node as the azimuth and the laser index of the parent node; inferring a radius of the predicted position to be a minimum radius value, wherein the minimum radius value is different than the radius of the parent node; and determining, based on the predicted position of the current node, the position of the current node.

Abstract: Systems and methods are provided for provided for automatic evaluation of sperm morphology. An image of a semen sample is obtained, and at least a portion of the image is provided to a convolutional neural network classifier. The convolutional neural network classifier evaluates the portion of the image to assign to the portion of the image a set of likelihoods that the portion of the image belongs to a plurality of output classes representing the morphology of sperm within the portion of the image. A metric is assigned to the semen sample based on the likelihoods assigned by the convolutional neural network.

Abstract: In an embodiment, a GAN model is trained based on an image dataset. A set of images of a first class is generated by the GAN model. Further, a first saliency map of a first generated image is determined by a neural network model. A second saliency map of a second image, belonging to the first class, from image dataset is determined by the neural network model. A first interpretability coefficient is determined, based on the first and second saliency maps. A first typicality score between the first generated image and a first set of images, belonging to the first class, from the image dataset, is determined. A second typicality score between a pair of generated images is determined. A second interpretability coefficient is determined basis the first and second typicality scores. An interpretability score associated with the GAN model is determined based on the first and second interpretability coefficients.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for selecting locations in an environment of a vehicle where objects are likely centered and determining properties of those objects. One of the methods includes receiving an input characterizing an environment external to a vehicle. For each of a plurality of locations in the environment, a respective first object score that represents a likelihood that a center of an object is located at the location is determined. Based on the first object scores, one or more locations from the plurality of locations are selected as locations in the environment at which respective objects are likely centered. Object properties of the objects that are likely centered at the selected locations are also determined.

Abstract: This application discloses a method for determining an implantation location of recommendation information performed at a computer device. The method includes: acquiring a target video; acquiring, according to a scene change status of the target video, a target video frame being used for location detection; performing image recognition on the target video frame to obtain masking information of the target video frame including a first region of an object of a target type in the target video frame; and determining an implantation location of recommendation information in the target video frame based on the first region. Image recognition processing is performed on the target video frame to obtain the masking information of the target video frame, so as to determine the implantation location in the target video frame on the basis of the first region corresponding to the object of the target type in the target video frame.

Abstract: A device for generating an image of an object by electromagnetic waves has a transmission device which is set up to radiate electromagnetic waves in the direction of the object, a receiving device which is set up to receive electromagnetic waves from the object, and a digital processing and control unit which is set up to generate image data of the object from the measured data. Here, the transmission device and the receiving device are arranged in at least one modular unit. The digital processing and control unit has an interface via which different modular units can be exchangeably coupled to the digital processing and control unit. Here, the interface is set up to transmit data to the modular unit and to receive from this, to transmit control signals to the transmission device and to the receiving device, and to supply the modular unit with energy.

Abstract: Systems, methods, tangible non-transitory computer-readable media, and devices for autonomous vehicle operation are provided. For example, a computing system can receive object data that includes portions of sensor data. The computing system can determine, in a first stage of a multiple stage classification using hardware components, one or more first stage characteristics of the portions of sensor data based on a first machine-learned model. In a second stage of the multiple stage classification, the computing system can determine second stage characteristics of the portions of sensor data based on a second machine-learned model. The computing system can generate an object output based on the first stage characteristics and the second stage characteristics. The object output can include indications associated with detection of objects in the portions of sensor data.

Abstract: In some aspects, the described systems and methods provide for validating performance of a model trained on a plurality of annotated pathology images. A pathology image is accessed. Frames are generated using the pathology image. Each frame in the set includes a distinct portion of the pathology image. Reference annotations are received from one or more users. The reference annotations describe at least one of a plurality of tissue or cellular characteristic categories for one or more frames in the set. Each frame in the set is processed using the trained model to generate model predictions. The model predictions describe at least one of the tissue or cellular characteristic categories for the processed frame. Performance of the trained model is validated based on determining a degree of association between the reference annotations and the model predictions for each frame and/or across all frames in the set of frames.

Abstract: Systems and methods are disclosed for predicting one or more medical conditions utilizing digital images and employing artificial intelligent algorithms. The system offers accurate predictions utilizing quantized pre-trained deep learning model. The pre-trained deep learning model is trained on data samples and later refined as the system processes more digital images or new medical conditions are incorporated. One pre-trained deep learning model is used to predict the probability of one or more medical conditions and identify locations in the digital image effected by the one or more medical conditions. Further, one pre-trained deep learning model utilizing additional data and plurality of digital images, forecasts rate of infection and spread of the medical condition over time.