Abstract: The disclosure relates to systems, methods and computer readable media for providing network-based identification, generation and management of a unique damage (finger) print of vehicle(s) by geodetic mapping of stable key points onto a ground truth 3D model of the vehicle, and vehicle parts—identified from the raw images using supervised and unsupervised machine learning. Specifically, the disclosure relates to System and methods for the generation of unique damage print on a vehicle that is obtained from captured images of the damaged vehicle, photogrammetrically localized to a specific vehicle part, and the computer programs enabling the method, the damage print configured to be used, for example, in fraud detection in insurance claims.

Abstract: A messaging system performs engagement analysis based on labels associated with content items produced by users of the messaging system. The messaging system is configured to process content items comprising images to identify elements in the images and determine labels for the images based on conditions indicating when to associate a label of the labels with an image of the images based on the elements in the image. The messaging system is further configured to associate the label with the content item, in response to determining to associate the label with the image, associating the label with the content item. The messaging system is further configured to determine engagement scores for the label based on interactions of users with the content items associated with label and adjust the engagement scores to determine trends in the labels to generate adjusted engagement scores.

Abstract: Aspects of the present invention relate to a machine learning system that is trained to identify forms, performing a method that includes receiving a form as an input image; identifying a field in the input image; identifying boundaries of the field; identifying locations of characters in the field; creating a two-dimensional space containing special characters; replacing the special characters with the characters in the field; identifying one or more keywords in the field based on identification of words and/or location of words; and responsive to an indication that the identifying one or more keywords yielded an incorrect result, updating the machine learning system. In another aspect, the machine learning system is used to identify forms, and can identify whether a form requires registration and, if registration is required, performing the registration.

Abstract: The technology described includes methods for pretraining a document encoder model based on multimodal self cross-attention. One method includes receiving image data that encodes a set of pretraining documents. A set of sentences is extracted from the image data. A bounding box for each sentence is generated. For each sentence, a set of predicted features is generated by using an encoder machine-learning model. The encoder model performs cross-attention between a set of masked-textual features for the sentence and a set of masked-visual features for the sentence. The set of masked-textual features is based on a masking function and the sentence. The set of masked-visual features is based on the masking function and the corresponding bounding box. A document-encoder model is pretrained based on the set of predicted features for each sentence and pretraining tasks. The pretraining tasks includes masked sentence modeling, visual contrastive learning, or visual-language alignment.

Abstract: A guidance processing apparatus (100) includes an information acquisition unit (101) that acquires a plurality of different pieces of guidance information on the basis of states of a plurality of people within one or more images, and a control unit (102) that performs control of a plurality of target devices present in different spaces or time division control of a target device so as to set a plurality of different states corresponding to the plurality of pieces of guidance information.

Abstract: A device may receive images of an object and information identifying the object, process the images using an artificial intelligence technique to identify parts of the object that are depicted in the images, and receive information identifying a location of damage on the object and information regarding the damage on the object. The device may process the information identifying the location of damage to identify a damaged part of the object, identify images depicting the damaged part, and identify, in the images, a location of the damaged part. The device may generate a first content item for display at the location of the damaged part in the images and generate a second content item for display with the images based on user interaction with the first content item, where the second content item includes information based on the information regarding the damage on the object.

Abstract: that is captured by one or more images (e.g., by a video or several pictures of the scene). The method analyzes a captured image to identify a unique predefined displayed content (e.g., a first pattern or a sub-pattern of the first pattern) in a first region of the image. When the predefined displayed content is identified in the captured image, the method analyzes a second region of the image that is associated with the first region, in order to determine whether a set of one or more objects is within the second region. The second region is the same as the first region in some embodiments, while in other embodiments it is a bounding region (e.g., bounding shape) that encompasses the first region and a third region around the first region. In yet other embodiments, the second region has another association with the first region. When the method determines that an object is within the second region, the method generates and sends a notification (e.g., an email, a text message, an in-app message, etc.

Abstract: A method and device for counting colonies in a sample in a medium. A bright field image is generated from light transmitted through the sample and the medium in a bright field configuration and collected by a detection unit, and a dark field image is generated from light scattered from the sample and the medium in a dark field configuration and collected by a detection unit. A fusion image is generated by a computer processing unit by combining two operand images, one an inverted image of the bright field image, and the other the dark field image. Colonies of the sample are counted by the computer processing unit based on processing of the fusion image. A single detection unit can generate the bright and dark field images, or the detection units can be separate. Light and dark diffusing surfaces can be separately retained or formed by a single diffuser.

Abstract: An information processing apparatus includes at least one processor, and the processor derives a property for at least one predetermined property item which is related to a structure of interest included in an image. The processor specifies a basis region serving as a basis for deriving the property related to the structure of interest for each property item and derives a basis image including the basis region.

Abstract: A guidance processing apparatus (100) includes an information acquisition unit (101) that acquires a plurality of different pieces of guidance information on the basis of states of a plurality of people within one or more images, and a control unit (102) that performs control of a plurality of target devices present in different spaces or time division control of a target device so as to set a plurality of different states corresponding to the plurality of pieces of guidance information.

Abstract: The disclosure relates to a method of encoding a digital image in order to compress same, the digital image being defined as a point cloud associating a set of N pixels, designated as vertices, to a scalar intensity value. The method aims at establishing triangulation vertices of the digital image and implements the principles of algorithmic topology.

Abstract: Provided are a method and apparatus for detecting a pattern image, and more particularly, a pattern image detection method and a pattern image detection apparatus for effectively detecting a pattern of a template image from a target image. The pattern image detection method and the pattern image detection apparatus can provide an effect of quickly and accurately detecting the pattern of the template image from the target image while reducing the amount of computation for detecting the pattern of the template image.

Abstract: The present disclosure is an apparatus and a method for providing a precise motion estimation learning model including, a database unit which stores a standard dataset labeled according to a first number of key points, an animation dataset labeled according to a second number of key points which is larger than the first number, and a photorealistic dataset having the second number of key points, a standard learning unit which learns the standard dataset for motion estimation to generate a standard learning model, an animation learning unit which retrains the animation dataset based on a weight of the standard learning model to generate an animation learning model, and a motion estimation learning unit which trains the photorealistic dataset based on the weight of the animation learning model to finely tune to generate a precise motion estimation learning model.

Abstract: The present invention discloses a decoupling divide-and-conquer facial nerve segmentation method and device. As for the characteristics of a small facial nerve structure and a low contrast, a facial nerve segmentation model including a feature extraction module, a rough segmentation module, and a fine segmentation module is constructed. The feature extraction module is configured to extract a low-level feature and a plurality of different- and high-level features. The rough segmentation module is configured to globally search the different- and high-level features for facial-nerve features and fuse them. The fine segmentation module is configured to decouple a fused feature to obtain a central body feature. After the central body feature is combined with the low-level feature to obtain an edge-detail feature, a space attention mechanism is used to extract attention features from the central body feature and the edge-detail feature, to obtain a facial nerve segmentation image.

Abstract: This disclosure is directed to, in part, mobile carts that are configured to determine their respective locations based on analysis of image data generated by cameras mounted to the respective carts. For instance, an example mobile cart may include at least one camera of a field-of-view directed substantially away from a cart and substantially towards an outward environment of the cart, such as toward an inventory location that houses one or more items. The mobile cart may generate image data representative of items housed at an inventory location adjacent to the cart and may use computer-vision techniques to analyze the image data and determine characteristics of these items. The mobile cart may then use this information to determine which section of multiple sections of a store in which the cart is currently located.

Abstract: A system receives identification of an item to be loaded into a vehicle and determines dimensions of the item. The system adds the item to a total load for an identified vehicle and, based on known dimensions of available space in the identified vehicle, determines one or more candidate placement options for each of one or more items comprising the total load, within available space within the vehicle, wherein candidate placement options for each item are determined so as not to overlap with placement options for another of the one or more items within a given configuration representing placement options for all of the one or more items comprising the total load. Additionally, the system presents a visual rendering of at least one given configuration that successfully allows the total load to fit within the available space.

Abstract: A system and method are disclosed for using a local neighborhood for determining similar targets in different documents or using implicit coordinates for obtaining a coordinate location of a target. The local neighborhood method may include identifying a first target in a first document; identifying one or more first elements within a first distance range from the first target; creating a first local neighborhood based on the identifying; determining that that first local neighborhood is similar to a third local neighborhood in a second document; and determining a second target in the second document that corresponds to the first target in the first document, based on the determining the similarity. The implicit coordinates method may include performing OCR on the first document to find the first target; and obtaining a first location of the first target by using at least one of OCR or element recognition.

Abstract: A device, system, and/or method may be used to provide an adaptive alignment. A first video data may be received. The first video data may comprise a first video frame and a second video frame. Sensor pose data may be determined. The pose data may be associated with the first video frame and the second video frame. An adjusted video frame may be determined based on the first video frame and a motion indicated by the pose data. A frame adjustment value may be determined by comparing a first pixel from the adjusted video frame to a second pixel from the second video frame. The frame adjustment value may correlate the pose data to the first video data. A second video data may be determined by applying the frame adjustment value.

Abstract: An apparatus includes at least one processing device configured to obtain input frames from a video. The at least one processing device is also configured to generate a forward flow from a first input frame to a second input frame and a backward flow from the second input frame to the first input frame. The at least one processing device is further configured to generate an occlusion map at an interpolated frame coordinate using the forward flow and the backward flow. The at least one processing device is also configured to generate a consistency map at the interpolated frame coordinate using the forward flow and the backward flow. In addition, the at least one processing device is configured to perform blending using the occlusion map and the consistency map to generate an interpolated frame at the interpolated frame coordinate.

Abstract: A method includes obtaining multiple video frames. The method also includes determining whether a bi-directional optical flow between the multiple video frames satisfies an image quality criterion for bi-directional consistency. The method further includes identifying a non-linear curve based on pixel coordinate values from at least two of the video frames. The at least two video frames include first and second video frames. The method also includes generating interpolated video frames between the first and second video frames by applying non-linear interpolation based on the non-linear curve. In addition, the method includes outputting the interpolated video frames for presentation.