Abstract: A video shooting method includes: enabling a camera application on a terminal and in a slow-motion video recording mode of the camera application, when a motion detection function of the camera application is enabled, starting to capture video frames after a video recording instruction input by a user is detected, recording a second video frame set at a first frame rate automatically when a motion of a human body in a captured first video frame set is a preset motion, and generating a target video, wherein the target video includes a slow-motion video clip, and wherein the slow-motion video clip includes video frames in the second video frame set.

Abstract: An image capture device with processors can execute a set of instructions stored in the system memory to: receive a plurality of time-sequenced images of the animal from the digital image sensor; determine, using an artificial intelligence module, a first birth in process from a first subset of images of the plurality of time-sequenced images of the animal; and determine an interval of time lapse between the first birth in process from the first subset of images of the plurality of time-sequenced images of the animal and a next birth in process from the next subset of images of the plurality of time-sequenced images of the animal, as determined using the artificial intelligence module, and when the interval of time lapse between the first birth in process and the next birth in process exceeds a predetermined amount trigger an action from an alert trigger.

Abstract: Systems and methods for identifying surgery tools of a robotic surgery system in surgery video based on robot kinematics are described. Kinematic data describing a position of a first robotic surgery arm including a surgery tool is used to determine a nominal position of the first robotic surgery arm. A search area within image data obtained during a robotic surgery from an endoscope attached to a second robotic surgery arm is determined based on the kinematic data and a nominal position of the surgery tool. The surgery tool is identified within the search area of the image data using an object detection technique.

Abstract: A method for recommending a product, includes: providing product information having a product description, a main image, and a plurality of alternate images; for each alternate image: performing saliency segmentation to obtain a segment and a background, defining a segment bounding box enclosing the segment, extracting image feature from the segment bounding box, extracting textual feature from the product description, extracting main image feature from the main image, matching the image feature of the segment bounding box to the textual feature and the main image feature, and determining the alternate image as a qualified alternate image if the image feature of the segment bounding box matches the textual feature and the main image feature; and when a number of the qualified alternate image equals to or is greater than a threshold number, recommending the product to customers.

Abstract: A drive assist apparatus includes a storage unit that stores a three-dimensional model indicating a moving region, an input unit that receives, from a sensor group installed in the moving region, first height information indicating a first height which is a height of the mobile object and second height information indicating a second height which is a height of an object that satisfies a predetermined distance criterion from the mobile object, an extraction unit that extracts, from the three-dimensional model, a first plan view based on the first height information and a second plan view based on the second height information, a generation unit that generates a combined map for two-dimensionally showing the moving region and assisting the driving of the mobile object by combining the first plan view and the second plan view, and an output unit that transmits a generated combined map to the mobile object.

Abstract: Embodiments of the present disclosure relate to a method, device and computer-readable storage medium for image processing. A method for image processing comprises: obtaining a first image and at least one second image, wherein each image of the at least one second image is a transformed image of the first image; and generating a first semantic segmentation label for the first image based on at least one semantic prediction result of the at least one second image. By using the first image and the first semantic segmentation label as training data to train an image semantic segmentation model, it helps to increase the precision of the trained image semantic segmentation model.

Abstract: A computer-implemented method is for classifying a region of interest in a medical image data set depicting a body part of a patient. The region of interest contains an object of potential pathological relevance. In an embodiment, the method includes a plurality of steps. One step is directed to generate a plurality of different representations of the region of interest. Another step is directed to determine, for each of the representations, a classification to generate a corresponding plurality of classifications. Thereby, each classification indicates a pathological relevance of the object in the respective representation. Yet, a further step is directed to calculate an ensemble classification for the region of interest based on the plurality of classifications.

Abstract: A calculation method of food volume, a calculation method of food calories, an electronic apparatus, an electronic device and a computer-readable storage medium are provided. The calculation method of food volume includes: acquiring actual size information of a stereo reference object; acquiring a reference image corresponding to the stereo reference object and a food image corresponding to food to be estimated; and acquiring volume information of the food to be estimated, according to the actual size information of the stereo reference object, reference region size information acquired based on the reference image, and food region size information acquired based on the food image. The calculation method of food volume reduces the complexity of operation and enhances the flexibility of application.

Abstract: An information generating apparatus includes: an information managing unit that associates and manages an image obtained by photographing skin of a person, marking information indicating one or more markings that one or more operators perform with respect to one or more characteristic parts of skin in the image, and amendment information indicating amendment that a specialist makes to the marking information; and a marking-UI processing unit that superimposes and displays the one or more markings, indicated by the marking information, and the amendment, indicated by the amendment information, on the image.

Abstract: A computer-implemented method of training an image analysis model. The image analysis model comprises a coupling layer that determines an output vector of discrete values from an input vector of discrete values. First, a machine learnable submodel of the coupling layer is trained to predict a second input part of the coupling layer from a first input part of the coupling layer. Next, the image analysis model is trained. This involves applying the coupling layer by applying the machine learnable submodel to the first input part to obtain a prediction of the second input part; and determining a second output part by applying an invertible mapping to the second input part defined by the prediction of the second input part. The mapping maps a predicted value of an element of the second input part to a fixed value independent from the predicted value.

Abstract: A registration means for storing an image of a product as a registration image in association with information representing the passing sequence that the product passed through an upstream side process; a management means for managing the matching sequence in a downstream side process; and a matching means for performing matching between an image of a product carried into the downstream side process and the registration image according to the matching sequence, are included. Each time the matching means succeeds in matching, the management means updates the matching sequence to sequence in which registration images not having succeeded in matching with any matching image are put in order on the basis of the passing sequence that the products passed through the upstream side process.

Abstract: The devices, systems, and methods can improve magnetic resonance imaging (MRI), MR spectroscopy (MRS), MR spectroscopic imaging (MRSI) measurement(s), thereby providing more reliable quantification. The method may include a method for correcting MR image(s)/spectrum. The method may include providing an inhomogeneity field/response map of a region of interest; and providing MR image(s)/spectrum of the region of interest. The method may include determining an intravoxel/voxel inhomogeneity correction coefficient for each voxel of at least one subregion of the region of the interest using the inhomogeneity field/response map. The method may include correcting each voxel of the MR image(s)/spectrum of the region of interest using the intravoxel/voxel inhomogeneity correction coefficient.

Abstract: Provided is an image analysis unit which analyzes a captured image of a camera mounted on a mobile device, executes object identification of an image, and sets a label as an identification result to an image region; a low-confidence region extraction unit which extracts a region with low confidence of object identification from an image analysis result; and a label updating unit which updates a label of the low-confidence region on the basis of information received via a communication unit. The label updating unit updates a label in a case where a matching rate between an object region analyzed from information received via the communication unit and the low-confidence region is equal to or greater than a specified threshold.

Abstract: An object amount calculation apparatus includes: a receiver configured to obtain a first three-dimensional model and a second three-dimensional model different from the first three-dimensional model, each of the first three-dimensional model and the second three-dimensional model representing a same space, each of the first three-dimensional model and the second three-dimensional model being constituted with regions having respective attributes; and a processor configured to: align the first three-dimensional model and the second three-dimensional model based on at least one attribute of the first three-dimensional model and the second three-dimensional model; calculate, for each of the attributes, a difference between the first three-dimensional model aligned and the second three-dimensional model aligned; and output (i) a total amount of differences corresponding to two or more attributes among the attributes and (ii) information on the two or more attributes.

Abstract: Facial recognition adversarial patch adjustment is performed by applying a patch to a digital image including image data representing a face to obtain a patched image, applying a face detector to the patched image to obtain a detection value representing a likelihood that the patched image includes image data representing facial features, applying a feature extractor to the patched image to obtain a feature vector, applying a similarity determiner to the feature vector and a target feature vector to obtain a similarity value representing a likelihood that the patched image includes image data representing a target face, calculating a loss based on the detection value and the similarity value, and adjusting the patch based on the loss.

Abstract: A label is attached to data corresponding to an activity included in time series data. A labeling system includes a processor and a machine learning model from which, in response to an input of time series data, a high-level label indicating a high-level activity included in the time series data and low-level labels indicating a plurality of low-level activities included in the high-level activity are outputted. The processor is configured to output a high-level label and a low-level label by using the machine learning model and cause the high-level label and the low-level label to be displayed, accept an input regarding a high-level label included in the time series data, train the machine learning model assuming that a high-level activity corresponding to the accepted high-level label is included in the time series data, and output a high-level label and a low-level label by using the trained machine learning model.

Abstract: One embodiment provides a method comprising acquiring 3D content comprising a 3D object in 3D space. The 3D object has object information indicative of a location of the 3D object in the 3D space. The method further comprises projecting the 3D object to a 2D object in 2D space based on the object information. The 2D object has one or more 2D vertices indicative of a location of the 2D object in the 2D space. The method further comprises determining one or more latent variables in the 2D space based on the object information and the one or more 2D vertices, and generating a 2D feature database including the one or more latent variables.

Abstract: An electronic device is provided that determines initial three-dimensional (3D) coordinates of a lighting device. The electronic device controls an emission of light from the lighting device based on control signals. The emitted light includes at least one of a pattern of alternating light pulses or a continuous light pulse. The electronic device controls a plurality of imaging devices to capture a first plurality of images that include information about the emitted light. Based on the determined initial 3D coordinates and the information about the emitted light included in the first plurality of images, the electronic device estimates a plurality of rotation values and a plurality of translation values of each imaging device. Based on the plurality of rotation values and the plurality of translation values, the electronic device applies a simultaneous localization and mapping process for each imaging device, for spatial synchronization of the plurality of imaging devices.

Abstract: A first object detection unit detects an object from a captured image and outputs object position information, in a non-high speed processing mode. A switching unit switches to a high-speed processing mode when the first object detection unit outputs the object position information. An image trimming unit extracts a trimmed image from the captured image based on the object position information output from the first object detection unit, in the high-speed processing mode. A second object detection unit detects an object from the trimmed image and outputs the object position information. A protrusion determination unit determines whether the object thus detected protrudes from the trimmed image. When it is determined that the object thus detected protrudes from the trimmed image, the switching unit switches to the non-high speed processing mode.

Abstract: The invention describes a generic framework for hand-eye calibration of camera-guided apparatuses, wherein the rigid 3D transformation between the apparatus and the camera must be determined. An example of such an apparatus is a camera-guided robot.