Abstract: The present invention concerns a method of segmenting a media item comprising media elements.

Abstract: The invention notably relates to a computer-implemented method for recognizing a three-dimensional modeled object from a two-dimensional image. The method comprises providing a first set of two-dimensional images rendered from three-dimensional modeled objects, each two-dimensional image of the first set being associated to a label; providing a second set of two-dimensional images not rendered from three-dimensional objects, each two-dimensional image of the second set being associated to a label; training a model on both first and second sets; providing a similarity metric; submitting a two-dimensional image depicting at least one object; and retrieving a three-dimensional object similar to the said at least one object of the two-dimensional image submitted by using the trained model and the similarity metric.

Abstract: An image inspection apparatus includes: a parameter setting unit that automatically sets a value of a first parameter among a plurality of image processing parameters; an image generation unit that generates a processed image with each of a plurality of second parameter candidate values obtained by changing a value of a second parameter; a display unit that displays a second parameter candidate list image group in which a plurality of processed images are listed; an image selection unit that receives a selection of any one of the processed images included in the second parameter candidate list image group displayed on the display unit; and a determination unit that outputs a determination result of the visual inspection. The parameter setting unit sets a second parameter candidate value corresponding to the processed image selected by the image selection unit as the second parameter.

Abstract: Embodiments of methods, apparatuses, and storage mediums associated with eye movement based knowledge demonstration, having a particular application to authentication, are disclosed. In embodiments, a computing device may determine whether a received input of a pattern of eye movements is consistent with an expected pattern of eye movements of a user when the user attempts to visually locate a piece of information embedded in a display. In embodiments, the expected pattern of eye movements may include patterns related to fixations and/or other statistical patterns, however, may not be limited to such patterns. In applications, determining consistency or correlation with the expected pattern of eye movements may identify the user by simultaneously verifying at least factors of authentication—that of biometric criteria related to a user's pattern of eye movements and a password or other information known to the user.

Abstract: A system for broad area geospatial object recognition, identification, classification, location and quantification, comprising an image manipulation module to create synthetically-generated images to imitate and augment an existing quantity of orthorectified geospatial images; together with a deep learning module and a convolutional neural network serving as an image analysis module, to analyze a large corpus of orthorectified geospatial images, identify and demarcate a searched object of interest from within the corpus, locate and quantify the identified or classified objects from the corpus of geospatial imagery available to the system. The system reports results in a requestor's preferred format.

Abstract: An image processing system transforms content images into the style of another reference style image. For example, the system applies a noise mask to generate noisy versions of the content images. The system then recomposes a content image in the style of a reference image by applying computer models to the noisy version of the content image, which reduces artifacts in the stylized image compared to that of a stylized image generated by applying the computer models to the original content image. When the content images are part of a video sequence, the image processing system may adjust the noise mask applied in a subsequent frame such that it tracks the movement of the client device from the current frame to the subsequent frame. This allows the system to reduce artifacts while stylizing the frames of the video in a consistent manner.

Abstract: Systems and methods directed toward combining visible light and infrared images can include processing visible light image data to determine an edge factor value for a plurality of visible light pixels corresponding to the strength of an edge at that location. The edge factor value can be determined using features from the visible light image data and an edge gain input, which may be adjustable by a user. The edge factor values are combined with an edge midscale value to create a first set of modified visible light image data including pixels emphasized based on the strength of the edge in the visible light image. The modified visible light image data is combined with infrared image data to create combined image data having contribution from the infrared image data and the edge factor values from the visible light image data.

Abstract: A supervision and penalty method and system for expressway emergency lane occupancy based on public supervision and reporting. The system includes a mobile terminal and a central computer. When a citizen notices a violation behavior of emergency lane occupancy of another vehicle, evidence is collected by photographing and images are uploaded to the central computer; meanwhile, user location information is automatically read and uploaded. The central computer acquires, through network transmission, image information and location information uploaded by a client, distinguishes whether it is a malicious reporting behavior and performs technical processing and license plate recognition on the uploaded images, checks the road section where the violation behavior occurs, and determines whether the violation behavior is an emergency lane violation behavior.

Abstract: Embodiments of the invention provide a method for scene understanding based on a sequence of image frames. The method comprises converting each pixel of each image frame to neural spikes, and extracting features from the sequence of image frames by processing neural spikes corresponding to pixels of the sequence of image frames. The method further comprises encoding the extracted features as neural spikes, and classifying the extracted features.

Abstract: A system and method for generate a mood log based on user images. In one embodiment, a system includes an image module that receives images taken by a user's mobile computing device and determines that a face of the user is included in the image, a mood module that determines a mood level of the user based on the face, and a log module that stores the mood level in a log of mood levels for the user.

Abstract: A processor of an orientation estimation device, as an orientation estimation method for estimating an orientation of a player existing in a sport video, receives reference orientation information which is information obtained based on information input by a user and which is information designating a joint position of a specific player existing in the sport video of an estimation target game and estimates an orientation of an estimation target player who is a player other than the specific player existing in a sport video of an estimation target using the reference orientation information.

Abstract: There are provided systems and method for three-dimensional (3D) eyeglasses modeling from two-dimensional (2D) images. A service provider, such as a payment provider, may access 2D images of a pair of eyeglasses, such as images from a front perspective and a side perspective of the eyeglasses. The contours of the eyeglasses may be extracted, such as the contours of the main features (e.g., the bridge, lens aperture, side and ear frame sections, and other important features). The contours may be processed to determine a curve that best fits the features. Additional features may be extracted, such as color, design, writing (e.g., a brand name on the frame), hinges, or other sub-feature of the frame. Utilizing this information, a 3D model of the pair of eyeglasses may be constructed by scaling the size of the lens curves, bridge, and/or arms and applying a texture.

Abstract: A system and method for extending bounds of straightened and enlarged images is described. A user interface of an image editing application exposes an image to a user for editing. The user positions the image in the user interface and the image editing application generates a frame in the user interface to identify boundaries of a final image in the user interface. The image editing application then automatically determines an empty area within the frame, applies a fill operation to the empty area within the frame, and updates the image in the user interface to reflect results of the fill operation.

Abstract: A method of generating adjusted image data to compensate for modality-induced distortion includes, at a processor: receiving a three-dimensional image captured with a first imaging modality and including (i) distorted surface image data depicting a surface of an object and (ii) distorted volume image data depicting a volume of the object; extracting the distorted surface image data from the three-dimensional image; receiving reference surface image data captured with a second imaging modality and depicting the surface of the object; determining a surface transformation for registering the distorted surface image data with the reference surface image data; determining a volume transformation informed by the surface transformation; generating an adjusted three-dimensional image by applying the volume transformation to the three-dimensional image; and storing the adjusted three-dimensional image in the memory.

Abstract: The present disclosure provides systems and methods to enable detection of linear objects such as utility poles in complex and heterogeneous outdoor surroundings. The methods deal with shape and orientation as prominent features of a pole model. Candidate trapeziums from 2D images of the face of the poles are extracted, some of which represent parts of the pole. To overcome the missed detection of certain parts due to problems of occlusion and diffusion into background, shape based techniques, that extrapolate and capture a longer trapezium representing the pole is implemented. The region growing stage or extrapolation is driven by orientation-based clustering of trapeziums. Context information is further used to identify objects of interest, by discarding false positives. Besides detecting poles of interest, the detected poles are further analyzed to identify damages, if any.

Abstract: A method for providing a contact card is provided. The method includes: acquiring a contact card of a contact person and acquiring contact information stored in a user terminal, the contact card including a first photo, and the contact information including a second photo; determining whether the first photo matches the second photo; and providing the contact card of the contact person to the user terminal, if the first photo matches the second photo.

Abstract: A system and method for determining a mood for a crowd is disclosed. In example embodiments, a method includes identifying an event that includes two or more attendees, receiving at least one indicator representing emotions of attendees, determining a numerical value for each of the indicators, and aggregating the numerical values to determine an aggregate mood of the attendees of the event.

Abstract: Systems and methods for determining dimensions of an object using a digital image. In particular, systems and methods for determining an actual dimension of a target object using a digital image of that object along with a reference object are disclosed. The digital image may be of a mirrored reflection of the reference object and the target object.

Abstract: Embodiments of the invention provide a method for scene understanding based on a sequence of image frames. The method comprises converting each pixel of each image frame to neural spikes, and extracting features from the sequence of image frames by processing neural spikes corresponding to pixels of the sequence of image frames. The method further comprises encoding the extracted features as neural spikes, and classifying the extracted features.

Abstract: Systems and methods are provided herein for identifying an identity of the person of interest in a live video, accessing his profile, and identifying closely related persons. Control circuitry may then identify a subset of the persons who are near a location where the live video is being filmed, compare the identities of each person of the subset to entries of a database of social media profiles, and identify respective social media components that correspond to each person of the subset. Finally, the control circuitry may determine whether a person of the subset has modified a social media component within a predefined period of time, and, in response to determining that the person of the subset has modified the respective social media component within the predefined period of time, provide information relating to the modification of the respective social media component.