Abstract: A computer-implemented method for an augmented-reality system is provided. The computer-implemented method comprises obtaining sensed data, representing an environment in which the AR system is located, determining that the AR system is in a location associated with a first authority characteristic, and controlling access to the sensed data for one or more applications operating in the AR system. Each of the one or more applications is associated with a respective authority characteristic. Controlling access to the sensed data for a said application is performed in dependence on the first authority characteristic and a respective authority characteristic associated with the said application. An AR system comprising one or more sensors, storage for storing sensed data, one or more application modules, and one or more processors arranged to perform the computer-implemented method is provided.

Abstract: A retina image template matching method is based on the registration and comparison between the images captured with portable low-cost fundus cameras (e.g., a consumer grade camera typically incorporated into a smartphone or tablet computer) and a baseline image. The method solves the challenges posed by registering small and low-quality retinal template images captured with such cameras. Our method combines dimension reduction methods with a mutual information (MI) based image registration technique. In particular, principle components analysis (PCA) and optionally block PCA are used as a dimension reduction method to localize the template image coarsely to the baseline image, then the resulting displacement parameters are used to initialize the MI metric optimization for registration of the template image with the closest region of the baseline image.

Abstract: A VR (Virtual Reality) simulator projects or displays a virtual space image on a screen installed at a position distant from a user in a real space and not integrally moving with the user. More specifically, the VR simulator acquires a real user position being a position of the user's head in the real space. The VR simulator acquires a virtual user position being a position in a virtual space corresponding to the real user position. Then, the VR simulator acquires the virtual space image by imaging the virtual space by using a camera placed at the virtual user position in the virtual space, based on virtual space configuration information indicating a configuration of the virtual space. Here, the VR simulator performs a lens shift process that shifts a lens of the camera such that the entire screen fits within a field of view of the camera.

Abstract: Iris image pick-up devices perform image pick-up of an iris of a moving subject. Images from the iris image pick-up devices are written to storage areas for transfer in a memory. A readout device reads out an image stored in the storage areas for transfer to a storage area for image processing at a frame interval corresponding to the movement speed of the subject. An image processing unit executes processing using the image read out to the storage area for image processing.

Abstract: Methods, systems, and computer-readable media for registering electromagnetic navigation data of a luminal network to a 3D model of the luminal network includes accessing a 3D model of a luminal network based on computed tomographic (CT) images of the luminal network, the 3D model corresponding to a CT coordinate space, selecting a plurality of reference points within the 3D model of the luminal network, accessing a plurality of survey points within the luminal network, the plurality of survey points being based on electromagnetic navigation data and corresponding to a body coordinate space, correlating the plurality of reference points and the plurality of survey points to determine pairs of correlated reference points and survey points; and deriving a transformation that maps the body coordinate space to the CT coordinate space based on the pairs of correlated reference points and survey points.

Abstract: A VR (Virtual Reality) simulator projects or displays a virtual space image on a screen installed at a position distant from a user in a real space and not integrally moving with the user. More specifically, the VR simulator acquires a real user position being a position of the user's head in the real space. The VR simulator acquires a virtual user position being a position in a virtual space corresponding to the real user position. Then, the VR simulator acquires the virtual space image by imaging the virtual space by using a camera placed at the virtual user position in the virtual space, based on virtual space configuration information indicating a configuration of the virtual space. Here, the VR simulator performs a lens shift process that shifts a lens of the camera such that the entire screen fits within a field of view of the camera.

Abstract: Embodiments of the present invention discloses an imaging calibration method and apparatus for a dual-light camera and a dual-light camera. A thermal imaging image including a first calibration frame image and a visible light image including a second calibration frame image are acquired. The first calibration frame image and the second calibration frame image are images of a same calibration frame respectively photographed by a thermal imaging camera and a visible light camera. First position information of the first calibration frame image in the thermal imaging image and second position information of the second calibration frame in the visible light image are acquired. Calibration parameters are generated according to the first position information and the second position information. A position of the thermal imaging image synthesized in the visible light image in the visible light image is adjusted according to the calibration parameters.

Abstract: A medical marking device, consisting of a radiotransparent plate having a plurality of radiopaque elements embedded therein in a predefined pattern. The device also has a sigmoid mounting arm having a first end connected to the radiotransparent plate and a second end containing one or more fastening receptacles.

Abstract: An appliance and method for automated visual inspection of at least two items from different categories, on a production line, include automatically switching between an item profile of a first of the at least two items for inspection; and an item profile of a second of the at least two items for inspection, based on detection of the first item in an image of the production line.

Abstract: The present disclosure relates to a technology field of artificial intelligence and provides a method for selecting image samples and related equipment. The method trains an instance segmentation model with first image samples and trains a score prediction model with third image samples. An information quantum score of second image samples is calculated through the score prediction model and feature vectors extracted. The second image samples are clustered according to the feature vectors of the second image samples and sample clusters of the second image samples are obtained. Target image samples are selected from the second image samples according to the information quantum score of the second image samples and the sample clusters. Target image samples from the image samples are selected for labelling, improving an accuracy of sample selection.

Abstract: Examples of methods for object deformation determination are described herein. In some examples, a method includes aligning a first bounding box of a three-dimensional (3D) object model with a second bounding box of a scan. In some examples, the method includes determining a deformation between the 3D object model and the scan based on the alignment.

Abstract: A dynamically controlled cooling device includes a stand having deployable legs for positioning the stand on a floor in a selectable position. A fan is coupled to the stand. A fan direction motor is coupled to the stand for selectively rotating the fan relative to the stand. A camera having a targeting function for identifying a person within a field of vision of the camera is coupled to the fan. A processor is operationally coupled to the camera and the fan direction motor to rotate the fan to point at the person.

Abstract: A disparity image fusion method for multiband stereo cameras belongs to the field of image processing and computer vision. The method obtains pixel disparity confidence information by using the intermediate output of binocular disparity estimation. The confidence information can be used to judge the disparity credibility of the position and assist disparity fusion. The confidence acquisition process makes full use of the intermediate output of calculation, and can be conveniently embedded into the traditional disparity estimation process, with high calculation efficiency and simple and easy operation. In the disparity image fusion method for multiband stereo cameras proposed by the method, the disparity diagrams participating in the fusion are obtained according to the binocular images of the corresponding bands, which makes full use of the information of each band and simultaneously avoiding introducing uncertainty and errors.

Abstract: A medical image processing apparatus includes processing circuitry configured to receive first volume data and second volume data; set a first symmetry plane for the first volume data and a second symmetry plane for the second volume data; and conduct a registration of the first volume data and the second volume data based on the first symmetry plane and the second symmetry plane.

Abstract: Disclosed biopsy markers are adapted to serve as localization markers during a surgical procedure. Adaptation includes incorporation of materials detectable under ultrasound during surgery, as well as features for co-registration with image guidance or other real-time imaging technologies during surgery. Such biopsy markers, when used as localization markers, improve patient comfort and reduce challenges in surgical coordination and surgery time. Additional disclosed biopsy markers are adapted to serve as monitoring and/or detection apparatuses, Localization of an implanted marker may be done with ultrasound technology. Ultrasound image data is analyzed to identify the implanted marker. A distance to the marker or a lesion may be determined and displayed. The determined distance may be a distance between the ultrasound probe and the marker or lesion, a distance between the marker or lesion and an incision instrument, and/or a distance between the ultrasound probe and the incision instrument.

Abstract: According to one or more embodiments, an image processing device includes a storage device and a processor. The storage device stores a plurality of time-series images. The processor extracts one or more feature points of a first image of the time-series images, sets a search range for one or more corresponding points of the feature points in a second image of the time-series images, searches for the one or more corresponding points in the search range of the second image, if there are a plurality of corresponding points searched, associates the searched corresponding points with the feature points based on a positional relationship between the corresponding points.

Abstract: An apparatus, system, method, and a control program on a recording medium are provided, each of which: acquires first point cloud data, and second point cloud data different from the first point cloud data; sets a plurality of search start positions each used for alignment form the first point cloud data to the second point cloud data; searches, for each of one or more of the plurality of search start positions, a solution candidate for coordinate transformation from the first point cloud data to the second point cloud data, to generate a plurality of solution candidates; and determines a final solution, from the plurality of solution candidates.

Abstract: Techniques for providing improved optical character recognition (OCR) for receipts are discussed herein. Some embodiments may provide for a system including one or more servers configured to perform receipt image cleanup, logo identification, and text extraction. The image cleanup may include transforming image data of the receipt by using image parameters values that optimize the logo identification, and performing logo identification using a comparison of the image data with training logos associated with merchants. When a merchant is identified, a second image clean up may be performed by using image parameter values optimized for text extraction. A receipt structure may be used to categorize the extracted text. Improved OCR accuracy is also achieved by applying on format rules of the receipt structure to the extracted text.

Abstract: An image processing system having at least one processor configured to: obtain a captured image of a document that includes a fixed part and an un-fixed part, the document being captured by an image reader or an image capture device; detect a feature part of the fixed part based on the captured image; obtain, in a case where the feature part is detected, a shaped image of the captured image such that a positional relationship of the feature part is aligned with a predetermined first positional relationship; and search for, in a case where the feature part is not detected, a substitute feature part of the fixed part based on the captured image, wherein in a case where the feature part is not detected, the at least one processor obtains the shaped image such that a position of the substitute feature part is aligned with a predetermined second position.

Abstract: The present disclosure relates to an image registration method and a model training method thereof. The image registration method comprises obtaining a reference image and a floating image to be registered, performing image preprocessing on the reference image and the floating image, performing non-rigid registration on the preprocessed reference image and floating image to obtain a registration result image, and outputting the registration result image. The image preprocessing comprises performing, on the reference image and the floating image, coarse-to-fine rigid registration based on iterative closest point registration and mutual information registration. The non-rigid registration uses a combination of a correlation coefficient and a mean squared error between the reference image and the registration result image as a loss function.

Abstract: The invention provides a system and related equipment for the precise measurement of the output characteristic of a light source, e.g., a dental light curing unit (LCU) or light for photodynamic therapy, using a light collector, a light detector, and a computer programmed to deliver the value of the output characteristic of the light source to the user. The systems allow for the determination of a proper exposure time or the selection of a light source as needed for a specific application.

Abstract: An apparatus for correcting a road region includes a processor configured to segment a road region extracted from an aerial image and representing a road into a plurality of partial road regions; associate, for each of the partial road regions, the partial road region with a road section existing at the location of the partial road region, the road section being indicated by map information indicating locations of respective road sections; and correct the road region so as to cover, for each of the partial road regions, the road section corresponding to the partial road region.

Abstract: Technologies for determining the accuracy of three-dimensional models include a device having circuitry to obtain two-dimensional images of an anatomical object (e.g., a bone of a human joint), to obtain a candidate three-dimensional model of the anatomical object, and to produce two-dimensional silhouettes of the candidate three-dimensional model. The circuitry is also to apply an edge detection algorithm to the two-dimensional images to produce corresponding edge images and to compare the two-dimensional silhouettes to the edge images to produce a score indicative of an accuracy of the candidate three-dimensional model.

Abstract: A method for decoding a 360-degree image includes: receiving a bitstream obtained by encoding a 360-degree image; generating a prediction image by making reference to syntax information obtained from the received bitstream; combining the generated prediction image with a residual image obtained by dequantizing and inverse-transforming the bitstream, so as to obtain a decoded image; and reconstructing the decoded image into a 360-degree image according to a projection format. Here, generating the prediction image includes: checking, from the syntax information, prediction mode accuracy for a current block to be decoded; determining whether the checked prediction mode accuracy corresponds to most probable mode (MPM) information obtained from the syntax information; and when the checked prediction mode accuracy does not correspond to the MPM information, reconfiguring the MPM information according to the prediction mode accuracy for the current block.

Abstract: Methods, apparatus and computer software products implement embodiments of the present invention that include applying energy to a probe that is in contact with tissue in a body cavity so as to ablate the tissue. While applying the energy, signals are received from a location transducer in the probe, which are indicative of a location of the probe in the cavity. The signals are processed so as to derive 3D location coordinate points corresponding to the location of the probe at a sequence of times during which the energy was applied. While applying the energy, a 3D representation of the body cavity is rendered to a display, and visual indicators are superimposed on the 3D representation, the visual indicators corresponding to the 3D location coordinate points at the sequence of times. Finally, a linear trace connecting the coordinate points in accordance with the sequence is superimposed on the 3D representation.

Abstract: Systems, instruments, and methods for medical treatment are disclosed. The methods comprise, by a computing device: receiving information identifying at least one first point on a body part shown in a medical image; overlaying a first mark on the medical image for the at least one first point; generating a spline based at least on the first mark; overlaying a second mark for the spline on the medical image; identifying a location of at least one second point on the body part shown in the medical image based on the first and second marks; overlaying a third mark for the at least one second point on the medical image; and using at least the third mark to facilitate the medical treatment of an individual whose body part is shown in the medical image.

Abstract: A reduced noise image can be formed from a set of images. One of the images of the set can be selected to be a reference image and other images of the set are transformed such that they are better aligned with the reference image. A measure of the alignment of each image with the reference image is determined. At least some of the transformed images can then be combined using weights which depend on the alignment of the transformed image with the reference image to thereby form the reduced noise image. By weighting the images according to their alignment with the reference image the effects of misalignment between the images in the combined image are reduced. Furthermore, motion correction may be applied to the reduced noise image.

Abstract: Embodiments of the present invention facilitate product defect detection. A computer-implemented method comprises: receiving, by a device operatively coupled to one or more processors, a template image of a normal product; generating, by the device, one or more geometric training parameters for transforming the template image; and transforming, by the device, the template image using the one or more geometric training parameters to generate a transformed image for training a data model, wherein the trained data model being used for aligning the template image and an image under inspection of a product.

Abstract: A method for rescue support involves distributing transponder devices to individuals in a geographic region and causing the transponder devices to be activated. If a large-scale emergency situation occurs, a fleet of unmanned aerial vehicles are operated within the geographic region to receive distress signals transmitted by the activated transponder devices. A computer system processes the distress signals to obtain detection data for the individuals, the detection data comprising a location of the respective transponder device and a health status of the respective individual. The computer system generates, based on the location and health status of the respective individual, a prioritization chart for the geographic region, the prioritization chart being indicative of one or more sub-regions to be prioritized for a rescue operation.

Abstract: Technology described herein includes a method that includes obtaining a set of color-coded sequences, each of which includes a sequence of colors. Each color-coded sequence has auto-correlation properties characterized by a merit factor larger than a first predetermined threshold, and cross-correlation properties among the color-coded sequences characterized by a demerit factor lower than a second predetermined threshold. A color-coded sequence is randomly selected from the set of color-coded sequences. A subject is illuminated in accordance with the sequence of colors in the selected color-coded sequence. A sequence of images of the subject are captured, and are temporally synchronized with illumination by the color-coded sequence. A filtered response image is generated from the sequence of images by a matched filtering process. Based on the filtered response image, it is determined that the subject is an alternative representation of a live person. In response, access to a secure system is prevented.

Abstract: Methods and systems for automatic estimation of object characteristics from a digital image are disclosed, including a method comprising sub-dividing into two or more segments a digital image comprising pixels and depicting an object of interest, wherein each segment comprises two or more pixels; assessing content depicted in one or more of the segments for a predetermined object characteristic using machine learning techniques comprising General Image Classification of the one or more segments using a convolutional neural network, wherein the General Image Classification comprises analyzing the segment as a whole and outputting a general classification for the segment as a whole as related to the one or more predetermined object characteristic; and determining a level of confidence of one or more of the segments having the one or more predetermined object characteristic based on the General Image Classification assessment.

Abstract: A method of performing personalized neuromodulation on a subject is provided. The method includes acquiring functional magnetic resonance imaging (fMRI) data of a brain of the subject. The method also includes calculating functional connectivity of the brain between a voxel in a subcortical region of the brain and a voxel in a cortical region of the brain, based on the fMRI data. The method also includes identifying a target location in the brain to be targeted by neuromodulation based on the calculated functional connectivity.

Abstract: The invention relates to a method for image integration (100), said method comprising the steps of acquiring at least two overlapping images (110) having different properties, and forming at least one integrated image (170) based on the overlapping images (110), wherein the step of forming at least one integrated image (170) comprises generating at least one score map (130) for each acquired overlapping image (110) relating to locally salient features of said overlapping image (110), forming an integration map (150) based on the at least one score map (130) of the at least two overlapping images (110), and integrating the at least two overlapping images (110) based on the integration map (150) to form an integrated image (170).

Abstract: A method for indicating a probing target for a fabricated electronic circuit including: generating an electronic, three-dimensional model based on manufacturing layout information of a fabricated circuit; obtaining, with a vision system, visual environment information for the fabricated circuit; scaling and orienting the three-dimensional model by a scaler and mapper based on the visual environment information; overlaying the three-dimensional model with the visual environment information to produce a correlated image; obtaining an identification of a desired network node of the fabricated circuit; and indicating a probing target, the probing target corresponding to the desired network node of the fabricated circuit.

Abstract: Described herein is a novel method and system for segmentation of the spine using 3D volumetric data. In embodiments, a method includes an extracting step, localization step, and segmentation step. The extracting step comprises detecting the spine centerline and the spine canal centerline. The localization step comprises localizing the vertebra and intervertebral disc centers. Background and foreground constraints are created for each vertebra digit. Segmentation is performed for each vertebra digit and based on the hard constraints.

Abstract: A compact microscope including an enclosure, a support element, a primary optical support element located within the enclosure and supported by the support element, at least one vibration isolating mount between the support element and the primary optical support element, a sample stage supported on the primary optical support element to support a sample, a return optical system to receive returned light from a sample and transmit returned light to a detection apparatus, wherein the return optical system is mounted on the primary optical support element, and wherein the compact microscope include a at least one of the following elements; a) an objective lens system, b) a temperature-control system, and c) the return optical system being operable to separate returned light into at least a first wavelength band and a second wavelength band.

Abstract: A CNN operates on the disparity or motion outputs of a block matching hardware module, such as a DMPAC module, to produce refined disparity or motion streams which improve operations in images having ambiguous regions. As the block matching hardware module provides most of the processing, the CNN can be small and thus able to operate in real time, in contrast to CNNs which are performing all of the processing. In one example, the CNN operation is performed only if the block hardware module output confidence level is below a predetermined amount. The CNN can have a number of different configurations and still be sufficiently small to operate in real time on conventional platforms.

Abstract: An appliance for concurrent automated visual inspection of at least two items comprising: at least two inspection assemblies each comprising a camera assembly, wherein each one of the at least two items is inspected by one of the at least two Inspection assemblies; a controller in data communication with the at least two inspection assemblies, wherein the controller is a computing device, wherein the appliance is adapted to be automatically configured for inspecting the at least two items using the at least two inspection assemblies.

Abstract: Examples disclosed herein may involve a computing system that is operable to (i) identify a source map and a target map for transferring map data, where the source map and the target map have different respective coordinate frames and respective coverage areas that at least partially overlap, (ii) select a real-world element for which to transfer previously-created map data from the source map to the target map, (iii) select a source image associated with the source map in which the selected real-world element appears and has been labeled, (iv) select a target image associated with the target map in which the selected real-world element appears, (v) derive a geometric relationship between the source image and the target image, and (vi) use the derived geometric relationship between the source image and the target image to determine a position of the real-world element within the respective coordinate frame of the target map.

Abstract: Methods and systems are provided which relate to the planning and delivery of radiation treatments by modalities which involve moving a radiation source along a trajectory relative to a subject while delivering radiation to the subject. An artificial intelligence (AI) agent trained using reinforcement learning (and/or some other suitable form of machine learning) is used to control the radiation delivery parameters in effort to achieve desired delivery of radiation therapy. In some embodiments, the AI agent selects suitable control steps (e.g. radiation delivery parameters for particular time steps), while accounting for patient motions, difference(s) in patient anatomical geometry and/or the like.

Abstract: A model-based method for quantifying a specimen. The method includes providing a specimen, capturing images of the specimen while illuminated by multiple spectra at different nominal wavelengths, and exposures, and classifying the specimen into various class types comprising one or more of serum or plasma portion, settled blood portion, gel separator (if used), air, tube, label, or cap; and quantifying of the specimen. Quantifying includes determining one or more of: a location of a liquid-air interface, a location of a serum-blood interface, a location of a serum-gel interface, a location of a blood-gel interface, a volume and/or a depth of the serum or plasma portion, or a volume and/or a depth of the settled blood portion. Quality check modules and specimen testing apparatus adapted to carry out the method are described, as are other aspects.

Abstract: A computing device spatially reconstructs a virtual feature surface in a mixed reality environment. The computing device detects addition of a raycast element to a virtual user space, maps multiple feature points detected from multiple video frames of a physical user space into a virtual user space, selecting at least three feature points from the multiple feature points that satisfy selection criteria applied in the virtual user space along a raycast axis of the raycast element in the virtual user space, and defines the virtual feature surface in the virtual user space using the at least three selected feature points. At least two of the at least three feature points are detected in different video frames.

Abstract: It is possible to determine a geometric transformation matrix representing geometric transformation between an input image and a template image with high precision. A geometric transformation matrix/inlier estimation section 32 determines a corresponding point group serving as inliers, and estimates the geometric transformation matrix representing the geometric transformation between the input image and the template image. A scatter degree estimation section 34 estimates scatter degree of the corresponding points based on the corresponding point group serving as inliers. A plane tracking convergence determination threshold calculation section 36 calculates a threshold used in convergence determination when iterative update of the geometric transformation matrix in a plane tracking section 38 is performed based on the estimated scatter degree.

Abstract: A method of processing image data in a connectionist network includes: determining, a plurality of offsets, each offset representing an individual location shift of an underlying one of the plurality of output picture elements, determining, from the plurality of offsets, a grid for sampling from the plurality of input picture elements, wherein the grid comprises a plurality of sampling locations, each sampling location being defined by means of a respective pair of one of the plurality of offsets and the underlying one of the plurality of output picture elements, sampling from the plurality of input picture elements in accordance with the grid, and transmitting, as output data for at least a subsequent one of the plurality of units of the connectionist network, a plurality of sampled picture elements resulting from the sampling, wherein the plurality of sampled picture elements form the plurality of output picture elements.

Abstract: Systems and methods for detecting patterns in data from a time-series are provided. In one implementation, a method for pattern detection includes obtaining data in a time-series and creating one-dimensional or multi-dimensional windows from the time-series data. The one-dimensional or multi-dimensional windows are created either independently or jointly with the time-series. The method also includes training a deep neural network with the one-dimensional or multi-dimensional windows utilizing historical and/or simulated data to provide a neural network model. Also, the method includes processing ongoing data with the neural network model to detect one or more patterns of a particular category in the ongoing data, and localizing the one or more patterns in time.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for image rendering. In one aspect, a method comprises receiving a plurality of observations characterizing a particular scene, each observation comprising an image of the particular scene and data identifying a location of a camera that captured the image. In another aspect, the method comprises receiving a plurality of observations characterizing a particular video, each observation comprising a video frame from the particular video and data identifying a time stamp of the video frame in the particular video. In yet another aspect, the method comprises receiving a plurality of observations characterizing a particular image, each observation comprising a crop of the particular image and data characterizing the crop of the particular image. The method processes each of the plurality of observations using an observation neural network to determine a numeric representation as output.

Abstract: A method of unsupervised learning of a metric representation and a corresponding system for a mobile device determines a metric position information for a mobile device from an environmental representation. The mobile device comprises at least one sensor for acquiring sensor data and an odometer system configured to acquire displacement data of the mobile device. An environmental representation is generated based on the acquired sensor data by applying an unsupervised learning algorithm. The mobile device moves along a trajectory and the displacement data and the sensor data are acquired while the mobile device is moving along the trajectory. A set of mapping parameters is calculated based on the environmental representation and the displacement data. A metric position estimation is determined based on a further environmental representation and the calculated set of mapping parameters.

Abstract: This disclosure describes interactive data processing systems configured to facilitate selection by a human associate of tentative results generated by an automated system from sensor data. In one implementation, an event may take place in a materials handling facility. The event may comprise a pick or place of an item from an inventory location, movement of a user, and so forth. The sensor data associated with the event is processed by an automated system to determine tentative results associated with the event. In some situations, an uncertainty may exist as to which of the tentative results accurately reflects the actual event. The system may then determine whether the event is to be merged with one or more temporally and spatially proximate events and, if so, the sensor data and tentative results for the merged event is sent to a human associate. The associate may select one of the tentative results.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for training an encoder neural network that is configured to process an input observation to generate a latent representation of the input observation. In one aspect, a method includes: obtaining a sequence of observations; for each observation in the sequence of observations, processing the observation using the encoder neural network to generate a latent representation of the observation; for each of one or more given observations in the sequence of observations: generating a context latent representation of the given observation; and generating, from the context latent representation of the given observation, a respective estimate of the latent representations of one or more particular observations that are after the given observation in the sequence of observations.

Abstract: The disclosure relates to PET imaging systems and methods. The systems may obtain a plurality of PET images of a subject and a CT image acquired by performing a spiral CT scan on the subject. Each gated PET image may include a plurality of sub-gated PET images. The CT image may include a plurality of sub-CT images each of which corresponds to one of the plurality of sub-gated PET images. The systems may determine a target motion vector field between a target physiological phase and a physiological phase of the CT image based on the plurality of sub-gated PET images and the plurality of sub-CT images. The systems may reconstruct an attenuation corrected PET image corresponding to the target physiological phase based on the target motion vector field, the CT image, and PET data used for the plurality of gated PET images reconstruction.