Abstract: A system may include a controller for receiving one or more images of a metrology target including periodic features with one or more known pitches, pre-processing the one or more images using a decomposition technique to generate one or more pre-processed images, and generating one or more metrology measurements for the metrology target based on the one or more pre-processed images. Pre-processing a particular image of the one or more images may include constructing one or more trajectory matrices from the particular image, generating reconstruction components associated with the particular image from the one or more trajectory matrices using the decomposition technique, and generating a particular one of the one or more pre-processed images by based on a subset of the reconstruction components including signals with at least one of the one or more known pitches.

Abstract: A method for estimating a pose of an object includes: receiving, by a processor, an observed image depicting the object from a viewpoint; computing, by the processor, an instance segmentation map identifying a class of the object depicted in the observed image; loading, by the processor, a 3-D model corresponding to the class of the object; computing, by the processor, a rendered image of the 3-D model in accordance with an initial pose estimate of the object and the viewpoint of the observed image; computing, by the processor, a plurality of dense image-to-object correspondences between the observed image of the object and the 3-D model based on the observed image and the rendered image; and computing, by the processor, the pose of the object based on the dense image-to-object correspondences.

Abstract: In a camera calibration apparatus (10), an external parameter calculation unit (13) calculates a rotation vector and a rotation matrix using the normalized start point coordinates, the normalized end point coordinates, and an “evaluation function” obtained by the normalized coordinate transformation unit (12) for each normal vector in the image plane and then outputs the calculated translation vector and rotation matrix. The normal vectors in the image plane respectively correspond to a plurality of normal vectors in a world coordinate space with respect to a reference plane in the world coordinate space. The evaluation function is for overdetermining a rotation parameter around a normal line with respect to the above reference plane in the world coordinate space.

Abstract: A system and method for adaptive color assignments to image labels during annotation of datasets includes preparing a dataset for image labeling by an annotator by: leveraging a global color analyzer to perform a global color distribution of a plurality of images to identify one or more overall colors present in the plurality of images, and a local color analyzer to perform a local color distribution for each image to identify one or more colors present in an area of interest of the image, and selecting a plurality of candidate colors to be used as image labels by the annotator, based on an output of the global color analyzer and an output of the local color analyzer.

Abstract: Systems, apparatuses, and methods for performing color channel correlation detection are disclosed. A compression engine performs a color channel transform on an original set of pixel data to generate a channel transformed set of pixel data. An analysis unit determines whether to compress the channel transformed set of pixel data or the original set of pixel data based on performing a comparison of the two sets of pixel data. In one scenario, the channel transformed set of pixel data is generated by calculating the difference between a first pixel component and a second pixel component for each pixel of the set of pixel data. The difference is then compared to the original first pixel component for each pixel. If the difference is less than or equal to the original for a threshold number of pixels, then the analysis unit decides to apply the color channel transform prior to compression.

Abstract: Systems and methods for measuring skin tone. A camera and computer may measure skin tone health of a face dataset and color quality of a single face image. The computer and camera may measure skin tone across wide sample of volunteers, express skin tone in CIELCH coordinates; and identify the range of observed lightness, chromaticity, and hue values in the population to determine a sample set. The computer and camera may be configured to match the image to a coordinate system of lightness, chromaticity, and hue in the sample set. If a match is not possible, the computer may be configured to select a coordinate closest to the measured lightness, chromaticity, and hue.

Abstract: The invention provides a method for managing image data in a motor vehicle lighting system, the lighting system including at least one lighting module intended to project light beams, the light beams being generated from data relating to the selection of at least one image, each image being respectively defined by a matrix including a plurality of horizontal or vertical rows of pixels, with each pixel having a numerical value related to a light intensity of the pixel. The method includes determining whether the pixel under analysis is considered to be a significant point of inflection of the image, so as to transmit it to at least one lighting module, so that it is able to project a resulting image.

Abstract: A three-dimensional data encoding method includes: (i) when a number of three-dimensional points included in point cloud data to be encoded is n that is greater than a predetermined number, n being an integer greater than or equal to 2, calculating an encoding coefficient by generating a hierarchical structure in which each of n pieces of attribute information on the three-dimensional points is sorted into one of a higher frequency component and a lower frequency component to be layered, and generating a bitstream including the encoding coefficient calculated in the calculating; and (ii) when a number of three-dimensional points included in the point cloud data is m that is smaller than or equal to the predetermined number, m being an integer greater than or equal to 1, generating a bitstream in accordance with m pieces of attribute information on the three-dimensional points without generating a hierarchy structure.

Abstract: Method, apparatus, and system for sampling-based dynamic mesh compression are provided. The process may include determining one or more sample positions associated with an input mesh based on one or more sampling rates, and an occupancy status associated respectively with each of the one or more sample positions indicating whether each of the one or more sample positions is within boundaries of one or more polygons defined by the input mesh is determined. The process may include generating a sample-based occupancy map based on the occupancy status associated respectively with each of the one or more sample positions.

Abstract: Provided are a method and electronic device for inferring an object in an image using a convolutional neural network (CNN) model. The method includes including: identifying a first region of interest in a first frame in the moving image, and a first object in the first region of interest, by providing the first frame to convolution layer groups sequentially connected in the CNN model, identifying a second region of interest in a second frame, the second region of interest corresponding to the first region of interest, and the second frame being after the first frame, providing the second region of interest to the CNN model, and obtaining first output data output from a first convolution layer group from among the convolution layer groups, and determining whether to identify a second object in the second region of interest by using a second convolution layer group, based on the first output data.

Abstract: A method for performing detection-based object searches includes receiving a user request indicating a region of interest, a timeframe of interest, or an object of interest. A signal is sent to cause execution of a query to identify object detections based on the user request. A signal representing at least one event identified in response to the query is received. For each event from the at least one event, a thumbnail image is identified based on the user request and using a ranking algorithm. A video frame identified based on the thumbnail image is received, and a video segment associated with the video frame is retrieved, A preview image clip that includes the video frame and the video segment is generated and displayed to a user associated with the user request.

Abstract: Systems, and method and computer readable media that store instructions for calculating signatures, utilizing signatures and the like, wherein for a low-power calculation of a signature, the method comprises: receiving or generating a media unit of multiple objects: processing the media unit by performing multiple iterations, determining a relevancy of the spanning elements of the iteration; completing the dimension expansion process by relevant spanning elements of the iteration and reducing a power consumption of irrelevant spanning; determining identifiers that are associated with significant portions of an output of the multiple iterations; and providing a signature that comprises the identifiers and represents the multiple objects.

Abstract: An information processing apparatus includes a processor configured to: read character string information written on a paper document; obtain, from an external system that performs processing using the read character string information, external system information indicating a use condition relating to use of character string information in the external system; and determine whether the read character string information matches the external system information obtained from the external system.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for detecting objects in images. One of the methods includes obtaining an input image; processing the input image to generate predicted part feature data, the predicted part feature data comprising, for each of a plurality of possible object parts: a part presence probability representing a likelihood that the possible object part is depicted in the input image, a predicted pose of the possible object part in the input image given that the possible object part is depicted in the input image, and an object part feature vector characterizing the depiction of the possible object part given that the possible object part is depicted in the input image; and processing the predicted part feature data for the plurality of possible object parts to generate an object detection output that identifies one or more objects depicted in the input image.

Abstract: A method of constructing a transformer model for answering questions about a video story according to an embodiment includes: extracting feature vectors related to each character of a video from video data including vision data and subtitle data and question data for video questions and answers, and generating an input embedding using the feature vectors related to the character; and training a transformer model using the input embedding.

Abstract: Systems and methods for identifying objects in shipping containers using radiographic images are described. More particularly, techniques for using artificial neural networks to analyze radiographic images of shipping containers and associated cargo manifests are discussed.

Abstract: The present disclosure relates to an information processing device, an information processing method, and a program for improving object recognition accuracy. A feature amount calculation unit of a user's car calculates a feature amount for an object recognition result in stages in each of a plurality of hierarchies, an acquisition unit acquires a feature amount for an object recognition result calculated by a feature amount calculation unit of another car, and a recognition unit performs the object recognition on the basis of the feature amount calculated by the user's car and the acquired feature amount calculated by the another car. The present disclosure can be applied to a mobile body.

Abstract: A Wildfire Early Detection System uses Aerial Video Clips of surveilled areas—obtained through a Network of Hardware components including UAVs and Tethered Aerostats equipped with a Camera and an AI-enabled Embedded Computer—and an Aerial Training Dataset, digitally combining variations of Smoke-Plume-Clips with variations of Background-Clips in nine different positions in the first frame of said Background-Clips with a relative size to the background calculated by perspective and programmed to follow the background to stay apparently “static” in the same place relative to the background for all the remaining frames of the Clip. A Computer-Vision Algorithm trained with that Aerial Training Dataset is used to recognize early fire Plumes of Smoke in those Aerial Video Clips with use of a Multiplication-free Neural Network for Wildfire detection, an AddNet based Discriminator CNN, a GANs used as both event detectors and smoke-plume scene synthesis and Block-based detection.

Abstract: A three-dimensional point cloud label learning and estimation device includes: a clustering unit that clusters a three-dimensional point cloud into clusters; a learning unit that makes a neural network learn to estimate a label corresponding to an object to which points contained in each of the clusters belong; and an estimation unit that estimates a label for the cluster using the neural network learned at the learning unit. In the three-dimensional point cloud label learning and estimation device, the neural network uses a total sum of sigmoid function values (sum of sigmoid) when performing feature extraction on the cluster.

Abstract: Systems, methods, and devices for detecting brake lights are disclosed herein. A system includes a mode component, a vehicle region component, and a classification component. The mode component is configured to select a night mode or day mode based on a pixel brightness in an image frame. The vehicle region component is configured to detect a region corresponding to a vehicle based on data from a range sensor when in a night mode or based on camera image data when in the day mode. The classification component is configured to classify a brake light of the vehicle as on or off based on image data in the region corresponding to the vehicle.

Abstract: An apparatus includes a processing node of a distributed computing platform. The processing node communicates with other processing nodes over one or more networks. The processing node may receive frames of point clouds at a processing node of a distributed computing platform, determine a subset of the frames as key frames based at least in part on distances travelled between captures of the respective frames, and allocate tasks of processing the key frames to processing subnodes based at least in part on estimated processing demands of the key frames and processing capabilities of each of the processing subnodes.

Abstract: A system includes: a named data networking (NDN) based Spark distributed computing network including a Spark distributed computing network including a master computer node and a plurality of slave computer nodes, and a named data networking (NDN) protocol installed on the Spark distributed computing network, and a coded distributed computing (CDC) target recognition model deployed on the NDN-based Spark distributed computing network. The NDN-based Spark distributed computing network is configured to: receive one or more batches of input images; generate a parity image from each batch of the input images; predict a label for each image of the batch of the input images; process the generated parity image; upon a label prediction of one image of the batch of the input images being unavailable, reconstruct the unavailable label prediction; and classify labels for the input images.

Abstract: Disclosed are methods and systems, among which is a system that includes a pattern-recognition processor, a central processing unit (CPU) coupled to the pattern-recognition processor via a pattern-recognition bus, and memory coupled to the CPU via a memory bus. In some embodiments, the pattern-recognition bus and the memory bus form about the same number of connections to the pattern-recognition processor and the memory, respectively.

Abstract: Methods and systems are described herein for hosting and arbitrating algorithms for the generation of structured frames of data from one or more sources of unstructured input frames. A plurality of frames may be received from a recording device and a plurality of object types to be recognized in the plurality of frames may be determined. A determination may be made of multiple machine learning models for recognizing the object types. The frames may be sequentially input into the machine learning models to obtain a plurality of sets of objects from the plurality of machine learning models and object indicators may be received from those machine learning models. A set of composite frames with the plurality of indicators corresponding to the plurality of objects may be generated, and an output stream may be generated including the set of composite frames to be played back in chronological order.

Abstract: The subject technology provides object recognition systems and methods that can be used to identify objects of interest in an image. An image such as live preview may be generated by a display component of the electronic device and an object of interest may be detected in the image. The detected object of interest may be classified using a classification model. Subsequent to classification, a confidence level in identifying the object of interest may be determined. In response to determining that the confidence level does not meet a confidence level threshold for identifying the object of interest, a request for a user input is generated. Based on the user input, the object of interest is identified using an object recognition model.

Abstract: An image processing apparatus includes a first detection unit configured to detect a person in an input image, a second detection unit configured to detect a specific object in the input image, a specification unit configured to specify a combination of the detected person and the detected object indicating an identical person, and a determination unit configured to determine a number of person in the input image, based on the detected person, the detected object, and the specified combination.

Abstract: An object locating method comprises deciding, for at least one frame of a video stream and based on at least one parameter representative of a change between a scene represented by the at least one frame and a scene represented by a reference frame of the video stream, whether determination of a location of at least one object (11, 13, 15, 17, 19) in the at least one frame is based on object detection applied to the at least one frame, or is based on a transformation of a location of the at least one object (11, 13, 15, 17, 19) detected in the reference frame.

Abstract: In one general aspect, a method can include capturing first features associated with a real-world physical area as a model and associating an AR object with a fixed location within the model. The method can include capturing second features associated with a real-world location corresponding with a portion of the real-world physical area. The method can include associating the captured second features with a location in the model, corresponding with the real-world location, as an AR anchor where the AR object is associated with the AR anchor.

Abstract: Disclosed is a method for the physical, in particular optical, detection of at least one usage object. The method includes the step of carrying out at least one physical detection process, for example by a user and/or an implementation device, in particular of at least one photograph, of the usage object, so that the usage object may be detected in such a way that an image of the usage object as detected during the detection process is shown at the same time as the database object shown on the screen in an identical manner or in a manner identical to scale, wherein as a result of the detection process, the usage object is associated with at least one usage object class, for example a vehicle type, by the processing unit and/or the CPU and/or the user.

Abstract: In order to improve a user's task of documenting items that are identified by the user while at a construction site, a computing system may be configured to (i) receive, from a first client station associated with a first user, media content that is descriptive of a given on-site item of a construction site, (ii) apply a speech recognition technique to transcribe audio data included in the media content into text that is descriptive of the given on-site item, (iii) generate a data record for the given on-site item, (iv) pre-populate at least a portion of the transcribed text into one or more fields of the data record for the given on-site item, and (v) cause a second client station to display the data record to a second user for review.

Abstract: A system includes a sensor capturing sensor data and an imager capturing imager data in combination with a computing device configured to time calibrate the sensor and the imager, detect an occurrence of a first event in the sensor data and extract a first time at which the first event occurred, and identify a first image in the image data having a capture time correlating to the first time at which the first event occurred in combination with retrieving the first image for insertion into a composite image including further image data combined with at least a portion of the first image depicting the first event.

Abstract: A computer-implemented method is provided for automated detection of action-activity and physical contact between sports events participants. Images of a field of play are captured by means of a video camera system producing at least one video stream which is digitally processed to identify human actions and any associated contact between contestants. The action-activity and contact detection happens with the usage of artificial neural network methods free of intermediate stages such as key-point and human pose-estimates.

Abstract: Certain aspects of the present disclosure provide a method of processing video data. In one example, the method includes receiving input video data; sampling a first subset of clips from the input video data; providing the first subset of clips to a first component of a machine learning model to generate first output; sampling a second subset of clips from the input video data, wherein the second subset of clips comprises fewer clips than the first subset of clips; providing the second subset of clips to a second component of the machine learning model to generate a second output; aggregating the first output from the first component of the machine learning model with the second output from the second component of the machine learning model to generate aggregated output; and determining a characteristic of the input video data based on the aggregated output.

Abstract: An image recognition approach employs both computer generated and manual image reviews to generate image tags characterizing an image. The computer generated and manual image reviews can be performed sequentially or in parallel. The generated image tags may be provided to a requester in real-time, be used to select an advertisement, and/or be used as the basis of an internet search. In some embodiments generated image tags are used as a basis for an upgraded image review. A confidence of a computer generated image review may be used to determine whether or not to perform a manual image review. Images and their associated image tags are optionally added to an image sequence.

Abstract: A plurality of similar video pairs may be determined based on one or more similarity information types. Each video pair of the plurality of similar video pairs may include a first respective video and a second respective video. For each video pair, one or more similar scene pairs may be determined. Each of the one or more similar scene pairs may include a respective first scene from the first respective video and a second respective scene from the second respective video. An encoder may be trained using a contrastive learning model that contrasts a plurality of similar scene pairs with a plurality of random scenes. The plurality of similar scene pairs may include the one or more scene pairs for each video pair. One or more scene features of one or more other scenes of one or more other videos may be determined using the encoder.

Abstract: In one embodiment, a method includes, by an electronic device, accessing activity data containing one or more non-image-based sensor signals from a first wearable device, where the activity data corresponds to an activity a user performs during a first timeframe, accessing from a first camera device, one or more cameras of the first camera device, where the video data corresponds to the first activity of the first user during the first timeframe, segmenting the activity data based on one or more features of the one or more non-image-based sensor signals to identify one or more segments of activity data corresponding to a second timeframe, classifying the one or more segments of the video data based on the one or more identified events associated with the first activity during the second timeframe, classifying the segments of the video data based on the one or more events during the second timeframe.

Abstract: Disclosed herein are system, method, and computer program product embodiments for camera surveillance planning and tracking using, for example, infrared technologies. An embodiment operates by receiving a camera specification input and placement input, calculating a surveillance coverage based on the camera specification and the placement input, displaying the surveillance coverage on a map, and identifying at least one stereo pair.

Abstract: A surveillance camera system according to an embodiment of the present disclosure comprises: multiple surveillance cameras for capturing images of different surveillance regions to surveil a protection object; an event management server which is connected to the surveillance cameras through a communication network and receives a first event signal or a second event signal from at least one of the surveillance cameras; and a manager terminal for receiving event information corresponding to the second event signal from the event management server when the second event signal is generated, wherein the first event signal corresponds to a signal generated when the protection object is detected by at least one of the multiple surveillance cameras, and the second event signal corresponds to a signal generated when the protection object is detected by none of the multiple surveillance cameras during a preconfigured reference time.

Abstract: A monitoring system can be configured to monitor activities or actions occurring in clinical settings, such as hospitals. The monitoring system can improve patient safety. The system can use visual and/or other tracking methods. The system can detect and/or identify people in a clinical setting. The system can also track activities of the people, for example, to improve adherence to hygiene protocols.

Abstract: A surveillance information generation apparatus (2000) includes a first surveillance image acquisition unit (2020), a second surveillance image acquisition unit (2040), and a generation unit (2060). The first surveillance image acquisition unit (2020) acquires a first surveillance image (12) generated by a fixed camera (10). The second surveillance image acquisition unit (2040) acquires a second surveillance image (22) generated by a moving camera (20). The generation unit (2060) generates surveillance information (30) relating to object surveillance, using the first surveillance image (12) and first surveillance information (14).

Abstract: Systems, methods, and other embodiments described herein relate to evaluating a perception network in relation to the accuracy of depth estimates and object detections. In one embodiment, a method includes segmenting range data associated with an image according to bounding boxes of objects identified in the image to produce masked data. The method includes comparing the masked data with corresponding depth estimates in the depth map according to an evaluation mask that correlates the depth estimates with the depth map. The method includes providing a metric that quantifies the comparing to assess a network that generated the depth map and the bounding boxes.

Abstract: Among other things, we describe systems and method for validating sensor calibration. For validating calibration of a system of sensors having several types of sensors, an object may be configured to have a substantially reflective portion such that the sensors can isolate the substantially reflective portion, and their sensor data can be compared to determine, if the detected locations of the substantially reflective portion by each sensor are aligned. For calibrating a system of sensors, an object having known calibration features can be used and detected by each sensor, and the detected data can be compared to known calibration data associated with the object to determine if each sensor is correctly calibrated.

Abstract: Systems, methods, and non-transitory computer-readable media can determine contextual information describing at least one physical structure corresponding to a location based at least in part on data captured by one or more sensors of a vehicle. A set of candidate interaction points for the at least one physical structure can be determined based at least in part on the determined contextual information describing the at least one physical structure corresponding to the location. The set of candidate interaction points can be filtered to identify one or more interaction points. An interaction point can be selected from the one or more interaction points to use for stopping the vehicle.

Abstract: The present disclosure is related in general to field of machine learning and image processing, that provides a method and system for detecting and classifying lanes. A lane detection and classification system receives an input image from a data source and segments the input image into plurality of segments using a trained semantic segmentation model. Further, one or more lane markings are detected in the segmented image and, lane pattern and lane colour of each of the one or more lane markings, wherein each lane marking is associated with a priority. Subsequently, a binary image comprising lane markings of ego lanes of the host vehicle is generated and coefficient values of the ego lanes of the host vehicle are determined based on the priority value associated with the lane markings of the ego lanes and current position of the host vehicle, using a trained Convolutional Neural Network (CNN) model.

Abstract: A vehicle control system includes an imaging device configured to capture an image of a travel route on which a vehicle is traveling, and a delimiting line estimating unit configured to estimate a position of a delimiting line on the travel route based on the image captured by the imaging device. The delimiting line estimating unit is configured to generate a first delimiting point sequence, generate a second delimiting point sequence that is offset rearward in a vehicle travel direction with respect to the first delimiting point sequence, generate a corrected delimiting point sequence based on the first delimiting point sequence and the second delimiting point sequence, and estimate the position of the delimiting line on the travel route based on the corrected delimiting point sequence.

Abstract: A method and an apparatus identifies an object. The method includes at a first time instant, acquiring raw data output by a camera installed on a vehicle, the output raw data comprises one of a continuous data output format for a continuous model or a discrete data output format for a discrete model; feeding the raw data to a discrete statistic model to form discrete statistic model format data for the first time instant, the discrete statistic model format data includes a continuous part having the continuous data output format and a discrete part having the discrete data output format; fusing the discrete statistic model format data for the first time instant with historical discrete statistic model format data to update the discrete statistic model format data for the first time instant; and identifying the object from the updated discrete statistic model format data for the first time instant, wherein the object includes a lanemarking characterized by a curve.

Abstract: An autonomous mobile platform is described. The autonomous mobile platform is configured for use in road monitoring, and comprising: a scanning module configured to be used for road scanning in order to enable identifying hazards associated with the road being scanned; a location finder, configured to determine locations of hazards identified at the road being scanned; and at least one transmitter comprised in a respective autonomous mobile platform, and configured to transmit data derived by the scanning module from the road scanning.

Abstract: A method for detecting groceries in corridor is provided, this method includes: obtaining an image collected from a corridor, performing pedestrian detection and grocery detection on the image collected from the corridor to obtain a pedestrian detection result and a grocery detection result; performing, if there is a pedestrian image in the pedestrian detection result, an image processing on the image collected from the corridor according to the pedestrian image; comparing the image that is collected from the corridor and has been processed with a preset corridor image to obtain an image similarity; generating, if there is a grocery image in the grocery detection result, or if the image similarity is less than or equal to a similarity threshold, a grocery cleaning instruction according to an identifier of corridor in the image collected from the corridor; and sending a grocery cleaning prompt according to the grocery cleaning instruction.

Abstract: An Enterobius vermicularis detection system is provided comprising a sample substrate, a sample disposed upon the sample substrate, and a personal computing device comprising an imaging device for acquiring one or more digital images of the sample substrate. The personal computing device may operate a digital interface allowing the acquired one or more digital images to be uploaded to one or more analysis computation servers. The one or more analysis computation servers may execute an analysis process upon the uploaded one or more digital images allowing determination of a presence of one or more Enterobius vermicularis eggs within the sample. The analysis process may output a first confidence interval value representing a likelihood of the presence of one or more Enterobius vermicularis eggs within the sample. The one or more analysis computation servers may communicate the first confidence interval value to the personal electronic device.

Abstract: A server and an electronic device for identifying a fake image are provided. The server includes a memory storing an artificial intelligence model trained to identify a fake image; and a processor connected to the memory, and configured to identify whether an image is the fake image by inputting the image to the artificial intelligence model, wherein the artificial intelligence model is a model trained based on an original image and a sample fake image, each including information about a landmark of a face area.