Abstract: A method of white point detection for an image is disclosed. The method includes determining a plurality of pixels of the image as a plurality of white point candidates; estimating a plurality of candidate confidences according to a plurality of brightness values corresponding to the plurality of white point candidates; and determining a plurality of white points of the image according to the plurality of candidate confidences.

Abstract: A hybrid-based image clustering method and a server for operating the same are provided. The method includes extracting a video object from each of a plurality of videos, calculating information about a similar key point between a predetermined reference target object and the extracted video object, clustering at least some of the plurality of videos into a first video group associated with the reference target object based on the information about the similar key point, and clustering at least some of the videos included in the first video group into a second video group based on at least one of video capture location information and place tag information of the videos included in the first video group.

Abstract: Systems and methods for analyzing, segmenting, and classifying multimedia material are disclosed herein. Embodiments include (i) receiving multimedia material for analysis, (ii) extracting elements from the multimedia material and forming objects comprising the elements; (iii) segmenting the multimedia material into segments, where individual segments include objects located within a threshold distance from each other; (iv) detecting objects within each segment; (v) associating attributes with the detected objects within the segments; (vi) annotating the segments by creating a relationship tree among the objects within each segment; and (vii) storing annotations of the segments for analysis.

Abstract: Certain embodiments involve generating recolored collages. For example, a method includes receiving a first image from and a second image. The method also includes grouping color hues within the first image into a first color hue cluster, and grouping color hues within the second image into a second color hue cluster. Further, the method includes determining that a first hue center of the first color hue cluster is within a distance range of a second hue center of the second color hue cluster. Additionally, the method includes generating a first recolored image by recoloring pixels of the first image having color hues outside the first color hue cluster and generating a second recolored image by recoloring pixels of the second image having color hues outside the second color hue cluster. Moreover, the method includes outputting a recolored collage including the first recolored image and the second recolored image.

Abstract: Systems and processes are disclosed for dynamically adjusting a speech trigger threshold, which can be used in triggering a virtual assistant. Audio input can be received via a microphone. The received audio input can be sampled, and a confidence level can be determined of whether the sampled audio input includes a portion of a spoken trigger. In response to the confidence level exceeding a threshold, a virtual assistant can be triggered to receive a user command from the audio input. The threshold can be dynamically adjusted in response to perceived events (e.g., events indicating a user may be more or less likely to initiate speech interactions, events indicating a trigger may be difficult to detect, events indicating a trigger was missed, etc.), thereby minimizing both missed triggers and false positive triggering events.

Abstract: A method for image processing, an apparatus having a processing circuitry and non-transitory computer readable medium are described. For example, the processing circuitry extracts, from a first image of a face, first values of feature points of a facial organ on the face. The feature points of the facial organ are indicative of characteristics of the facial organ. Then, the processing circuitry calculates first distances between the feature points of the facial organ based on the first values of the feature points. Further, the processing circuitry calculates feature parameters of the facial organ according to the first distances between the feature points. Then, the processing circuitry detects a facial action of the face when the feature parameters satisfy a preset condition associated with the facial action.

Abstract: The invention relates to the processing operation of interpolating the colours of a Bayer mosaic image sensor. A first elementary matrix filter, which is a bilinear interpolation filter, of size m×m, m being an odd number larger than or equal to 3, a low-pass matrix filter of size n×n, n being an odd number larger than or equal to 3, and a high-pass matrix filter, complementary to the low-pass filter, of size n×n, are defined. The first matrix filter is convoluted with the low-pass filter, resulting in a low-frequency interpolation filter of size (m+n?1)×(m+n?1), and the first matrix filter is convoluted with the high-pass filter, resulting in a high-frequency interpolation filter of size (m+n?1)×(m+n?1). The matrix of digital signals arising from the pixels is filtered separately, using the pixels of each colour, by the low-frequency interpolation filter. The complete matrix of signals is filtered using the high-frequency interpolation filter.

Abstract: A method and an apparatus for processing a high dynamic range (HDR) image, and a terminal device to improve quantization quality, where the method includes obtaining brightness information of an image, processing the brightness information to obtain processed image information, quantizing the processed image information to obtain quantized image information, and encoding the quantized image information to obtain encoded image information.

Abstract: Embodiments of the present disclosure provide a fingerprint registration method and device. The method includes: obtaining a fingerprint to be registered and a target fingerprint, extracting fingerprint features from the fingerprint to be registered and the target fingerprint, in which the fingerprint features include a ridge feature and a phase feature, performing a rough registration using ridge features of the two fingerprints, and then obtaining phase difference information of phase features of the two fingerprints in a phase overlap region, adjusting the rough registration result according to the phase difference information to obtain a final registration result of the two fingerprints. In some embodiments, it does not simply depend on the ridge feature to perform the fingerprint registration and add the phase feature on the basis of the ridge feature, and perform twice registration, thereby improving the accuracy of fingerprint registration.

Abstract: A system and processing methods for configuring and utilizing a convolutional neural network (CNN) for plant disease recognition are disclosed. In some embodiments, the system is programmed to collect photos of infected plants or leaves where regions showing symptoms of infecting diseases are marked. Each photo may have multiple marked regions. Depending on how the symptoms are sized or clustered, one marked region may include only one lesion caused by one disease, while another may include multiple, closely-spaced lesions caused by one disease. The system is programmed to determine anchor boxes having distinct aspect ratios from these marked regions for each convolutional layer of a single shot multibox detector (SSD). For certain types of plants, common diseases lead to relatively many aspect ratios, some having relatively extreme values. The system is programmed to then train the SSD using the marked regions and the anchor boxes and apply the SSD to new photos to identify diseased plants.

Abstract: An image processing apparatus capable of performing pixel synthesis processing at high speed. Synthesis processing of a first object and a second object is performed for an area where the first object and the second object overlap each other. A first image formation processor stores first pixel data, which corresponds to the area, of pixel data of the first object, in a first communication memory and stores second pixel data, which corresponds to the area, of pixel data of the second object, in a second communication memory. A second image formation processor acquires the first and second pixel data from the first and second communication memories, respectively, and performs synthesis processing of the acquired first and second pixel data.

Abstract: Embodiments of the present disclosure provide an image data processing method, an image data processing apparatus, an electronic device and a storage medium. The image data processing method includes: during a preset first collecting time period, collecting regularly a plurality of first user images corresponding to a target user, and extracting first part image data in each of the plurality of first user images; recording position information of the first part image data in each of the plurality of first user images on a display interface; if it is determined that the target user is at a preset stable state, performing a statistical processing to obtain a total movement times corresponding to the first part image data within a preset statistical time period; and if the total movement times reaches a preset threshold, performing an image processing on second part image data on a current display interface.

Abstract: Techniques for using deep learning to facilitate patch-based image inpainting are described. In an example, a computer system hosts a neural network trained to generate, from an image, code vectors including features learned by the neural network and descriptive of patches. The image is received and contains a region of interest (e.g., a hole missing content). The computer system inputs it to the network and, in response, receives the code vectors. Each code vector is associated with a pixel in the image. Rather than comparing RGB values between patches, the computer system compares the code vector of a pixel inside the region to code vectors of pixels outside the region to find the best match based on a feature similarity measure (e.g., a cosine similarity). The pixel value of the pixel inside the region is set based on the pixel value of the matched pixel outside this region.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for performing dynamic, stroke-based alignment of touch displays. In one aspect, a method includes obtaining a candidate transcription that an automated speech recognizer generates for an utterance, determining a particular context associated with the utterance, determining that a particular n-gram that is included in the candidate transcription is included among a set of undesirable n-grams that is associated with the context, adjusting a speech recognition confidence score associated with the transcription based on determining that the particular n-gram that is included in the candidate transcription is included among the set of undesirable n-grams that is associated with the context, and determining whether to provide the candidate transcription for output based at least on the adjusted speech recognition confidence score.

Abstract: A magnetic resonance imaging apparatus according to an embodiment includes a display, an input interface, and processing circuitry. The display displays at least a locator image and a reference image. The input interface sets a region of interest on the locator image displayed on the display. The processing circuitry scans a subject to obtain three dimensional data, generates a locator image from the three dimensional data and displaying the locator image on the display, generates a reference image corresponding to the location of the region of interest and displaying the reference image on the display, and makes, when a size or position of the region of interest on one of the locator image and the reference image is changed by the input interface, adjustments to correspondingly change the display magnification or position of the other one of the locator image and the reference image.

Abstract: The present invention is to reduce the time required to detect an object after completion of the rotation of a head or a body of a robot. A robot 100 includes a camera 111, and a control unit 127 which determines an overlapping area between a first image captured with the camera 111 at first timing and a second image captured with the camera 111 at second timing later than the first timing to detect an object included in an area of the second image other than the determined overlapping area.

Abstract: Systems and methods for providing location based voice recognition include receiving, through a first microphone, an audio signal from a first user that includes an audio command requesting a service that requires user authorization before access to at least a portion of the service is granted. The user authorization is based on voice recognition (e.g., voice authentication and/or voice identification) of the audio signal. The source location of the audio signal is determined and a user location of the first user is determined. If the source location of the audio signal correlates with the user location, voice recognition on the audio signal may be performed. The first user may be authorized to access the service based on the voice recognition performed on the audio signal.

Abstract: An image processing apparatus includes a processor and a memory having stored thereon instructions executable by the processor to cause the image processing apparatus to perform: calculating, with respect to each of a plurality of captured images obtained by successively capturing a subject, an evaluation index as an index as to whether a capturing condition is appropriate or not for each of a plurality of partial images of the each of the plurality of captured images corresponding to different areas of the subject; selecting the plurality of partial images corresponding to the different areas of the subject from the plurality of captured images based on the evaluation indices of the partial images; and synthesizing the plurality of selected partial images in positions corresponding to partial images in a reference image obtained by capturing the subject.

Abstract: Embodiments of the present invention are generally directed to estimating capacity usage of a container. In an embodiment, the present invention is a method of estimating a fullness of a container. The method includes: mounting an image capture apparatus proximate a container-loading area, the image capture apparatus operable to capture three-dimensional images; capturing, via the image capture apparatus, a three-dimensional image representative of a three-dimensional formation, the three-dimensional image having a plurality of points with three-dimensional point data including depth data; generating a histogram of the depth data from the three-dimensional image; and estimating the fullness of the container based at least in part on the histogram.

Abstract: The present disclosure provides systems and methods for generating an electrophysiological map of a geometric structure. The system includes a computer-based model construction system configured to acquire electrical information at a plurality of diagnostic landmark points, assign a color value, based on the acquired electrical information, to each of the diagnostic landmark points, create a first 3D texture region storing floats for a weighted physiological metric, create a second 3D texture region storing floats for a total weight, for each diagnostic landmark point, additively blend the color value of the diagnostic landmark point into voxels of the first 3D texture region that are within a predetermined distance, normalize the colored voxels using the second 3D texture region to generate a normalized 3D texture map, generate the electrophysiological map from the normalized 3D texture map and a surface of the geometric structure, and display the generated electrophysiological map.