Abstract: Systems and techniques are described for colorizing vector images. Color raster images that correspond to the vector images are used to perform a color mapping to colorize the vector images. The vector images are represented using non-overlapping planar arrangements of faces of the vector images, so that the color mapping may be performed with respect to the faces. The faces may be processed in parallel, to further facilitate a speed and scalability of the describes processes and results.

Abstract: According to one embodiment, a learning device includes an acquirer, a generator, a first identifier, an extractor, a second identifier, and a learning-processor. The acquirer acquires real data. The generator generates pseudo data of the real data using a first neural network. The first identifier identifies whether input data is the real data or the pseudo data. The extractor extracts features of data from the input data. The second identifier identifies whether the features extracted by the extractor are features of the real data or features of the pseudo data. The learning-processor learns the first neural network such that the pseudo data and the real data are not discriminated on the basis of an identification result of each identifier.

Abstract: A method for proactively creating an image product includes capturing an image of an object in a first environment by a device, storing a library of personalized products each characterized by a product type, automatically recognizing the object in the image as having a product type associated with the library of personalized products, automatically creating a design for the personalized product of the product type using personalized content, automatically displaying the design of the personalized product of the product type incorporating the selected photo in the first environment on the device, and manufacturing a physical product based on the design of the personalized product.

Abstract: Digital image completion by learning generation and patch matching jointly is described. Initially, a digital image having at least one hole is received. This holey digital image is provided as input to an image completer formed with a dual-stage framework that combines a coarse image neural network and an image refinement network. The coarse image neural network generates a coarse prediction of imagery for filling the holes of the holey digital image. The image refinement network receives the coarse prediction as input, refines the coarse prediction, and outputs a filled digital image having refined imagery that fills these holes. The image refinement network generates refined imagery using a patch matching technique, which includes leveraging information corresponding to patches of known pixels for filtering patches generated based on the coarse prediction. Based on this, the image completer outputs the filled digital image with the refined imagery.

Abstract: A method includes receiving media data corresponding to a media content item. The method includes analyzing the media data to determine characteristics of the media content item based on first visual information of the media content item. The method includes analyzing the characteristics of the media content item based on a media popularity data structure to generate a prediction of whether the media content item is likely to exceed a popularity threshold within a particular period of time. The method further includes determining a location to store the media data based on the prediction.

Abstract: A color information display device includes an extracting unit that extracts a reference color combination corresponding to reference color vision characteristics; an output unit that outputs the reference color combination as it is or after conversion in accordance with a degree of sharing of an impression of the reference color combination between the reference color vision characteristics and other color vision characteristics; and an assigning unit that assigns information indicative of the degree of sharing of the impression to the output reference color combination.

Abstract: Disclosed are an apparatus for measuring distance using two-step tracking and a method thereof. More particularly, the apparatus includes a bit generator configured to generate coarse bits for first tracking from sensing data to measure distance to a target and fine bits for second tracking, corresponding to the coarse bits, to measure precise distance to the target; a coarse bit processor configured to receive the coarse bits, address the received coarse bits in any one histogram bin, which corresponds to the received coarse bits, among a plurality of histogram bins, and output coarse bits, which correspond to any one histogram bin exceeding a preset threshold among the histogram bins, as reference coarse bits; and a fine bit processor configured to output fine bits corresponding to the reference coarse bits.

Abstract: A method for training a learning-based medical scanner including (a) obtaining training data from demonstrations of scanning sequences, and (b) learning the medical scanner's control policies using deep reinforcement learning framework based on the training data.

Abstract: A method for proactively creating an image product includes capturing an image of an object in a first environment by a device, storing a library of personalized products each characterized by a product type, automatically recognizing the object in the image as having a product type associated with the library of personalized products, automatically creating a design for the personalized product of the product type using personalized content, automatically displaying the design of the personalized product of the product type incorporating the selected photo in the first environment on the device, and manufacturing a physical product based on the design of the personalized product.

Abstract: A three-dimensional image reconstruction method, in which a cross section preparation step includes: a specimen preparation step of attaching, on a surface of a specimen, a grid-like mark member in which rectangular openings are two-dimensionally arranged, and disposing the grid-like mark member under a shielding member so that a side of each of the rectangular openings of the grid-like mark member forms a 45-degree or 90-degree angle with respect to a direction in which a linear end edge of the shielding member extends; and a processing position determination step of adjusting relative positions of the shielding member and the grid-like mark member by using the grid-like mark member as an index of a processing position.

Abstract: A computer-implemented method for determining and evaluating an objective tumor response to an anti-cancer therapy using cross-sectional images can include receiving cross-sectional images of digital medical image data and identifying target lesions within the cross-sectional images. For each of the target lesions, a target lesion type and anatomical location is identified, a segmenting tool is activated for segmenting the target lesions into regions of interest, lesion metrics are automatically extracted from the regions of interest according to tumor response criteria, and conformity of target lesion identification is monitored using rules associated with the tumor response criteria, prompting a user to address any nonconforming target lesion. The method also includes receiving a presence/absence of metastases, determining changes in lesions metrics, and deriving an objective tumor response based on the tumor response criteria.

Abstract: In real biometric systems, false match rates and false non-match rates of 0% do not exist. There is always some probability that a purported match is false, and that a genuine match is not identified. The performance of biometric systems is often expressed in part in terms of their false match rate and false non-match rate, with the equal error rate being when the two are equal. There is a tradeoff between the FMR and FNMR in biometric systems which can be adjusted by changing a matching threshold. This matching threshold can be automatically, dynamically and/or user adjusted so that a biometric system of interest can achieve a desired FMR and FNMR.

Abstract: An image retrieval assist device that is communicably connected to an investigation assist device for recording captured images of cameras installed at a plurality of intersections respectively in correlation with camera information and intersection information. The image retrieval assist device has a processor, a communication unit; and a storage that stores an extraction condition of an image for each type of event that has occurred at the intersection, the image indicating a situation at a time of occurrence of the event. The processor accepts an input of event information including the type of event. The processor generates an image retrieval key including the extraction condition according to the input event information. The communication unit transmits the generated image retrieval key to the investigation assist device.

Abstract: A method and apparatus for receiving a depth frame from a depth sensor oriented towards an open end of a shipping container, the depth frame comprising a plurality of grid elements that each have a respective depth value, identifying one or more occlusions in the depth frame, correcting the one or more occlusions in the depth frame using one or more temporally proximate depth frames, and outputting the corrected depth frame for fullness estimation.

Abstract: Provided are neural network-based image processing method and apparatus, and a computer-readable storage medium. The image processing method includes: inputting an image into an optimized neural network; extracting, by the optimized neural network, image features of the image; and outputting the image features, wherein the optimized neural network is obtained by performing a first optimization process on at least one sub-layer in a pre-trained initial neural network, each sub-layer of the at least one sub-layer includes a convolutional layer, and the first optimization process comprises: for each sub-layer of the at least one sub-layer, determining one or more channels to be removed from a filter of the convolutional layer and removing said one or more channels, and optimizing parameters of remaining channels in the filter of the convolutional layer, so that error of output features of each optimized sub-layer is minimized.

Abstract: The present invention is directed towards providing automated workflows for the identification of a reading order from text segments extracted from a document. Ordering the text segments is based on trained natural language models. In some embodiments, the workflows are enabled to perform a method for identifying a sequence associated with a portable document. The methods includes iteratively generating a probabilistic language model, receiving the portable document, and selectively extracting features (such as but not limited to text segments) from the document. The method may generate pairs of features (or feature pair from the extracted features). The method may further generate a score for each of the pairs based on the probabilistic language model and determine an order to features based on the scores. The method may provide the extracted features in the determined order.

Abstract: A structural analysis computing device for determining structural characteristics of an object pictured in a three-dimensional (3D) image may be provided. The structural analysis computing device may include a memory, a user interface, an object sensor configured to capture the 3D image of the object, and at least one processor in communication with the memory and the object sensor. The processor may be configured to access the 3D image including the object, automatically determine a first plurality of measurements of the object from the 3D image, and display the 3D image on the user interface. The processor may be further configured to generate a data file including the 3D image and the first plurality of measurements, and store the data file within the memory. The processor may also be configured to transmit the data file to an insurance server computing device for generation of an associated insurance claim form.

Abstract: A system determines spatial locations of pixels of an image. The system includes a processor configured to: receive location data from devices located within a hotspot; generate a density map for the hotspot including density pixels associated with spatial locations defined by the location data, each density pixel having a value indicating an amount of location data received from an associated spatial location; match the density pixels of the density map to at least a portion of the pixels of the image; and determine spatial locations of the at least a portion of the pixels of the image based on the spatial locations of the matching density pixels of the density map. In some embodiments, the image and density map are converted to edge maps, and a convolution is applied to the edge maps to match the density map to the pixels of the image.

Abstract: The image coding method includes: determining a context for a current block to be processed, from among a plurality of contexts; and performing arithmetic coding on the control parameter for the current block to generate a bitstream corresponding to the current block, wherein in the determining: the context is determined under a condition that control parameters of neighboring blocks of the current block are used, when the signal type is a first type, the neighboring blocks being a left block and an upper block of the current block; and the context is determined under a condition that the control parameter of the upper block is not used, when the signal type is a second type, and the second type is one of “mvd_l0” and “mvd_l1”.

Abstract: An image processing apparatus includes a search range providing circuit, a searching circuit and an encoder. The search range providing circuit obtains N number of predicted motion vectors for a target image block, and accordingly sets N number of corresponding search ranges in a reference frame, where N is an integer greater than 1. The searching circuit performs a motion compensation searching process individually on the N search ranges to accordingly determine a motion vector for the target image block. The encoder encodes the target image block according to the motion vector.