Abstract: A deep neural network (DNN) can be trained based on a first training dataset that includes first images including annotated first objects. The DNN can be tested based on the first training dataset to determine first object predictions including first uncertainties. The DNN can be tested by inputting a second training dataset and outputting first object predictions including second uncertainties, wherein the second training dataset includes second images including unannotated second objects. A subset of images included in the second training dataset can be selected based on the second uncertainties, The second objects in the selected subset of images included in the second training dataset can be annotated. The DNN can be trained based on the selected subset of images included in the second training dataset including the annotated second objects.

Abstract: In a method for determining if a test data set is anomalous in a deep neural network that has been trained with a plurality of training data sets resulting in back propagated training gradients having statistical measures thereof, the test data set is forward propagated through the deep neural network so as to generate test data intended labels including at least original data, prediction labels, and segmentation maps. The test data intended labels are back propagated through the deep neural network so as to generate a test data back propagated gradient. If the test data back propagated gradient differs from one of the statistical measures of the back propagated training gradients by a predetermined amount, then an indication that the test data set is anomalous is generated. The statistical measures of the back propagated training gradient include a quantity including an average of all the back propagated training gradients.

Abstract: Neural networks and learning algorithms can use a variance of gradients to provide a heuristic understanding of the model. The variance of gradients can be used in active learning techniques to train a neural network. Techniques include receiving a dataset with a vector. The dataset can be annotated and a loss calculated. The loss value can be used to update the neural network through backpropagation. An updated dataset can be used to calculate additional losses. The loss values can be added to a pool of gradients. A variance of gradients can be calculated from the pool of gradient vectors. The variance of gradients can be used to update a neural network.

Abstract: In a method of generating a neural network used to detect a feature of medical significance from a body image data input, test data images (110) are divided into patches (120). Each patch is labelled as either corresponding to the feature (124) or not corresponding to the feature (122). One trained fully connected layer (136) in a pretrained general purpose convolutional neural network (130) is replaced with a new fully connected layer. The pretrained convolutional neural network (130) is retrained with the set of labelled patches (126) to generate a feature-specific convolutional neural network that includes at least one feature-specific fully connected layer (152) that maps the body image data to the feature of medical significance when the feature of medical significance is present in the body image data input.

Abstract: A system for monitoring ocular movement can comprise a housing, a plurality of light sources, at least one imager, and a controller. The housing can define a cavity configured to allow each eye of a patient to view an interior region of the housing. The plurality of light sources can be oriented within the interior region of the housing. The at least one imager can be oriented to capture an image of an eye of a patient during an evaluation. The at least one controller can comprise at least one processor and a non-transitory computer readable medium storing instructions. The instructions can be executed by the processor and cause the controller to receive image data from the at least one imager and illuminate the plurality of light sources in a predetermined and reconfigurable sequence.

Abstract: A computing device, programmed to: acquire a color image and transform the color image into a color-component map. The computer can be further programmed to process the color-component map to detect a traffic sign by determining spatial coincidence and determining temporal consistency between the color-component map and the traffic sign.

Abstract: A system for monitoring ocular movement can comprise a housing, a plurality of light sources, at least one imager, and a controller. The housing can define a cavity configured to allow each eye of a patient to view an interior region of the housing. The plurality of light sources can be oriented within the interior region of the housing. The at least one imager can be oriented to capture an image of an eye of a patient during an evaluation. The at least one controller can comprise at least one processor and a non-transitory computer readable medium storing instructions. The instructions can be executed by the processor and cause the controller to receive image data from the at least one imager and illuminate the plurality of light sources in a predetermined and reconfigurable sequence.

Abstract: A plurality of data streams are received from a vehicle, at least one of the data streams including multimedia data. A prioritization of the data streams is performed according to one or more factors. At least one of adjusting at least one of the data streams and preventing transmission of at least one of the data streams according to the prioritization is performed.

Abstract: A computing device, programmed to: acquire a color image and transform the color image into a color-component map. The computer can be further programmed to process the color-component map to detect a traffic sign by determining spatial coincidence and determining temporal consistency between the color-component map and the traffic sign.

Abstract: A plurality of data streams are received from a vehicle, at least one of the data streams including multimedia data. A prioritization of the data streams is performed according to one or more factors. At least one of adjusting at least one of the data streams and preventing transmission of at least one of the data streams according to the prioritization is performed.

Abstract: Methods for hierarchical hole-filling and depth adaptive hierarchical hole-filling and error correcting in 2D images, 3D images, and 3D wrapped images are provided. Hierarchical hole-filling can comprise reducing an image that contains holes, expanding the reduced image, and filling the holes in the image with data obtained from the expanded image. Depth adaptive hierarchical hole-filling can comprise preprocessing the depth map of a 3D wrapped image that contains holes, reducing the preprocessed image, expanding the reduced image, and filling the holes in the 3D wrapped image with data obtained from the expanded image. These methods are can efficiently reduce errors in images and produce 3D images from a 2D images and/or depth map information.

Abstract: A coding method for an ordered series of quantized transform coefficients of a block of image data, including a context adaptive position coding process to encode the position of clusters of non-zero-valued coefficients, e.g., a multidimensional position coder that uses one of a plurality of code mappings selected according to at least one criterion including at least one context-based criterion, and an amplitude encoding process to encode any amplitudes remaining to be coded, the amplitude coding using one or a plurality of amplitude code mappings selected according to at least one criterion, including a context-based criterion. A context-based selection criterion means a criterion that during encoding is known or derivable from one or more previously encoded items of information. Also a coding apparatus, a decoding apparatus, a computer readable medium configured with instructions that when executed implement a coding method, and another medium for a decoding method.

Abstract: A coding method for an ordered series of quantized transform coefficients of a block of image data, including a context adaptive position coding process to encode the position of clusters of non-zero-valued coefficients, e.g., a multidimensional position coder that uses one of a plurality of code mappings selected according to at least one criterion including at least one context-based criterion, and an amplitude encoding process to encode any amplitudes remaining to be coded, the amplitude coding using one or a plurality of amplitude code mappings selected according to at least one criterion, including a context-based criterion. A context-based selection criterion means a criterion that during encoding is known or derivable from one or more previously encoded items of information. Also a coding apparatus, a decoding apparatus, a computer readable medium configured with instructions that when executed implement a coding method, and another medium for a decoding method.

Abstract: A coding method for an ordered series of quantized transform coefficients of a block of image data, including a context adaptive position coding process to encode the position of clusters of non-zero-valued coefficients, e.g., a multidimensional position coder that uses one of a plurality of code mappings selected according to at least one criterion including at least one context-based criterion, and an amplitude encoding process to encode any amplitudes remaining to be coded, the amplitude coding using one or a plurality of amplitude code mappings selected according to at least one criterion, including a context-based criterion. A context-based selection criterion is meant a criterion that during encoding is known or derivable from one or more previously encoded items of information. Also a coding apparatus, a decoding apparatus, a computer readable medium configured with instructions that when executed implement a coding method, and another medium for a decoding method.

Abstract: Methods for hierarchical hole-filling and depth adaptive hierarchical hole-filling and error correcting in 2D images, 3D images, and 3D wrapped images are provided. Hierarchical hole-filling can comprise reducing an image that contains holes, expanding the reduced image, and filling the holes in the image with data obtained from the expanded image. Depth adaptive hierarchical hole-filling can comprise preprocessing the depth map of a 3D wrapped image that contains holes, reducing the preprocessed image, expanding the reduced image, and filling the holes in the 3D wrapped image with data obtained from the expanded image. These methods are can efficiently reduce errors in images and produce 3D images from a 2D images and/or depth map information.

Abstract: A coding method, apparatus, and medium with software encoded thereon to implement a coding method. The coding method includes encoding cluster of consecutive non-zero-valued coefficients, the encoding of a cluster including jointly encoding joint events that each are defined by at least two parameters: the number of zero-valued coefficients preceding the cluster, and the number of non-zero-valued coefficients in the cluster. The encoding of the cluster also includes encoding a parameter indicative of the number of amplitude-1 trailing non-zero-valued coefficients in the cluster, in one version with the parameter indicative of the number of trailing amplitude-1 coefficients part of the joint events such that the coding is according to a 3-dimensional joint variable length coding table. The method further includes encoding the amplitudes of the non-zero-valued coefficients that are not encoded by the joint encoding, e.g., encoding the amplitudes of the other than the trailing amplitude-1 coefficients.

Abstract: A coding method, apparatus, and medium with software encoded thereon to implement a coding method. The coding method includes jointly encoding joint events that each are defined by a cluster of consecutive non-zero-valued coefficients, each joint event defined by three parameters: the number of zero-valued coefficients preceding the cluster, the number of non-zero-valued coefficients in the cluster, and an indication of which trailing coefficients up to a maximum number of M trailing coefficients have amplitude greater than 1, with the coding using a 3-dimensional joint VLC table. The method further includes encoding the amplitude of the non-zero-valued trailing coefficients that have amplitude greater than 1 encoding the amplitude of any remaining non-zero-valued coefficients in the clusters that have more than M non-zero-valued coefficients.

Abstract: A coding method for an ordered series of quantized transform coefficients of a block of image data, including a context adaptive position coding process to encode the position of clusters of non-zero-valued coefficients, e.g., a multidimensional position coder that uses one of a plurality of code mappings selected according to at least one criterion including at least one context-based criterion, and an amplitude encoding process to encode any amplitudes remaining to be coded, the amplitude coding using one or a plurality of amplitude code mappings selected according to at least one criterion, including a context-based criterion. A context-based selection criterion is meant a criterion that during encoding is known or derivable from one or more previously encoded items of information. Also a coding apparatus, a decoding apparatus, a computer readable medium configured with instructions that when executed implement a coding method, and another medium for a decoding method.

Abstract: A coding method, apparatus, and medium with software encoded thereon to implement a coding method. The coding method includes jointly encoding joint events that each are defined by a cluster of consecutive non-zero-valued coefficients, each joint event defined by three parameters: the number of zero-valued coefficients preceding the cluster, the number of non-zero-valued coefficients in the cluster, and an indication of which trailing coefficients up to a maximum number of M trailing coefficients have amplitude greater than 1, with the coding using a 3-dimensional joint VLC table. The method further includes encoding the amplitude of the non-zero-valued trailing coefficients that have amplitude greater than 1 encoding the amplitude of any remaining non-zero-valued coefficients in the clusters that have more than M non-zero-valued coefficients.

Abstract: A coding method, apparatus, and medium with software encoded thereon to implement a coding method. The coding method includes encoding cluster of consecutive non-zero-valued coefficients, the encoding of a cluster including jointly encoding joint events that each are defined by at least two parameters: the number of zero-valued coefficients preceding the cluster, and the number of non-zero-valued coefficients in the cluster. The encoding of the cluster also includes encoding a parameter indicative of the number of amplitude-1 trailing non-zero-valued coefficients in the cluster, in one version with the parameter indicative of the number of trailing amplitude-1 coefficients part of the joint events such that the coding is according to a 3-dimensional joint variable length coding table. The method further includes encoding the amplitudes of the non-zero-valued coefficients that are not encoded by the joint encoding, e.g., encoding the amplitudes of the other than the trailing amplitude-1 coefficients.