Abstract: A processor of an object detection device is configured to search for a pair of coordinates associated with a disparity approximately equal to a target disparity satisfying a predetermined condition in a second direction of a disparity map, update a first pair of coordinates to the found pair of coordinates, and calculate a height of an object corresponding to the target disparity on the basis of the first pair of coordinates. In the disparity map, a disparity acquired from a captured image is associated with a pair of two-dimensional coordinates formed by a first direction and the second direction intersecting the first direction. When a second pair of coordinates associated with a disparity approximately equal to the target disparity is present within a predetermined interval from the first pair of coordinates, the processor is configured to update the first pair of coordinates to the second pair of coordinates.

Abstract: A system for measuring distances within a surgical site includes a camera positionable to capture 3D image data corresponding to a treatment site. Using the image data, the system determines the relative 3D positions of identified measurement points at the treatment site captured in the images, and it estimates or determines the distance between the measurement points. Output is generated communicating the measured distances to the user. The measurement taken follows the 3D topography of the tissue positioned between the measurement points.

Abstract: The embodiments of the present disclosure disclose an object detection method and system, and a non-transitory computer-readable medium. In the object detection method, multiple candidate bounding boxes of an interest object in a current image frame are obtained. Based on a determined bounding box of the interest object in a previous image frame, the multiple candidate bounding boxes are filtered to obtain a determined bounding box of the interest object in the current image frame.

Abstract: Methods and systems are disclosed for generating training data for a machine learning model for better performance of the model. A source image is selected from an image database, along with a target image. An image segmenter is utilized with the source image to generate a source image segmentation mask having a foreground region and a background region. The same is performed with the target image to generate a target image segmentation mask having a foreground region and a background region. Foregrounds and backgrounds of the source image and target image are determined based on the masks. The target image foreground is removed from the target image, and the source image foreground is inserted into the target image to create an augmented image having the source image foreground and the target image background. The training data for the machine learning model is updated to include this augmented image.

Abstract: Methods are provided for the automated detection of micro-objects in a microfluidic device. In addition, methods are provided for repositioning micro-objects in a microfluidic device. In addition, methods are provided for separating micro-objects in a spatial region of the microfluidic device.

Abstract: The disclosed system may include a housing dimensioned to secure various components including at least one physical processor and various sensors. The system may also include a camera mounted to the housing, as well as physical memory with computer-executable instructions that, when executed by the physical processor, cause the physical processor to: acquire images of a surrounding environment using the camera mounted to the housing, identify features of the surrounding environment from the acquired images, generate a map using the features identified from the acquired images, access sensor data generated by the sensors, and determine a current pose of the system in the surrounding environment based on the features in the generated map and the accessed sensor data. Various other methods, apparatuses, and computer-readable media are also disclosed.

Abstract: The present disclosure relates to systems, methods, and non-transitory computer-readable media that generates object masks for digital objects portrayed in digital images utilizing a detection-masking neural network pipeline. In particular, in one or more embodiments, the disclosed systems utilize detection heads of a neural network to detect digital objects portrayed within a digital image. In some cases, each detection head is associated with one or more digital object classes that are not associated with the other detection heads. Further, in some cases, the detection heads implement multi-scale synchronized batch normalization to normalize feature maps across various feature levels. The disclosed systems further utilize a masking head of the neural network to generate one or more object masks for the detected digital objects. In some cases, the disclosed systems utilize post-processing techniques to filter out low-quality masks.

Abstract: An article image data acquirer acquires article image data as a target of optical character recognition (OCR). An inference result data generator generates first inference result data and second inference result data by inputting the article image data as the target of the OCR into a trained model. An inference result data outputter outputs the first inference result data and the second inference result data. An image filter generator generates a first image filter based on the first inference result data and a second image filter based on the second inference result data. An image filter outputter outputs the first image filter and the second image filter.

Abstract: A liveness detection method and apparatus, and a facial verification method and apparatus are disclosed. The liveness detection method includes detecting a face region in an input image, measuring characteristic information of the face region, adjusting the measured characteristic information in response to the characteristic information not satisfying a condition, and performing a liveness detection on the face region with the adjusted characteristic information upon the measured characteristic information not satisfying the condition.

Abstract: According to one embodiment, a learning apparatus includes a processor. The processor acquires first training data. The processor inputs the first training data to a model, and generate a plurality of estimation vectors that are a processing result of the model. The processor generates an estimation distribution from the estimation vectors. The processor calculates a distribution loss between the estimation distribution and a target distribution that is a target in an inference using the model. The processor updates parameters of the model, based on the distribution loss.

Abstract: A system and method for lateral vehicle detection is disclosed. A particular embodiment can be configured to: receive lateral image data from at least one laterally-facing camera associated with an autonomous vehicle; warp the lateral image data based on a line parallel to a side of the autonomous vehicle; perform object extraction on the warped lateral image data to identify extracted objects in the warped lateral image data; and apply bounding boxes around the extracted objects.

Abstract: A method of building an abnormality quantifier comprising: generating at least one selected first dataset comprising measurements of a normal population or sample and at least one second selected dataset comprising measurements of an abnormal population or sample; generating an image or map by imagizing the datasets; identifying a normality zone within the image or map using the first dataset; identifying an abnormality zone within the image or map using the second dataset; determining a definition of abnormality based on a comparison of the normality zone and the abnormality zone; receiving or accessing at least one third dataset comprising measurements of a both known normal and abnormal population or sample; testing the performance of the initially defined abnormality against one or more preset performance criteria; and outputting an abnormality quantifier when optimal performance has been reached.

Abstract: An information processing apparatus (2) includes: a reducing unit (211) that generates a second image (IMG2) by reducing a first image (MG1) in which a target object is included; a first extracting unit (212) that extracts, as a first key point, a key point (KP) of the target object from the second image; a setting unit (213) that sets a target area (TA) that designates a part of the first image based on the first key point; and a second extracting unit (214) that extracts, as a second key point, a key point of the target object from a target image (IMG1_TA) of the first image that is included in the target area.

Abstract: A target detection system suitable for an embedded device, comprising an embedded device (5) and a server (6); target detection logic (5.1) running in the embedded device (5) is composed of a multi-layer shared base network, a private base network, and a detection module; a parameter of the shared base network directly comes from an output of an upper layer; and an image is processed by the shared base network and the private base network to obtain a feature map, and after being processed by the detection module, a result merging module merges and outputs a target detection result. The target detection system further comprises an online model self-calibration system. After collecting a sample, the embedded device (5) irregularly uploads the sample to the server (6), and after labeling the sample by means of automatic and manual methods, the server (6) trains a model and updates same to the embedded device (5).

Abstract: The invention provides a camera extrinsic parameter calibration method, an image stitching method and an apparatus thereof. The calibration method can include: acquiring a plurality of images captured by the camera based on a plurality of sets of extrinsic parameters to be calibrated, wherein each of the plurality of sets of extrinsic parameters to be calibrated include a horizontal rotation angle and a vertical rotation angle; obtaining at least one matched image pair based on feature point matching among the plurality of images, wherein each matched image pair includes matched feature point pairs with a quantity satisfying a threshold condition; and calibrating the plurality of sets of extrinsic parameters to be calibrated according to a first difference representation between coordinate representations of each matched feature point pair of each matched image pair in a reference coordinate system.

Abstract: A vehicle parking finder support system, method and computer program product for determining if a vehicle is at a reference parking location. The vehicle parking finder support system includes at least a first camera configured to provide images of the surroundings of the vehicle; a memory; a processing circuitry operatively connected to the at least first camera and the memory, configured to cause the vehicle parking finder support system to: obtain, by the at least first camera, at least first image data of the surroundings of the vehicle; obtain, from the memory, at least first reference image data; and determine if the vehicle is at the reference parking location using the at least first reference image data and the obtained at least first image data.

Abstract: The disclosure purposes to provide an optical coherence tomography (OCT)-based system for diagnosing a high risk lesion such as a vulnerable atheromatous plaque by using an artificial intelligence model through deep learning. A deep learning-based diagnostic method of diagnosing a high risk lesion of a coronary artery includes: acquiring an OCT image of a coronary artery lesion of a patient; extracting a first feature of a thin cap from the OCT image; setting a region of interest included in the OCT image on a basis of the first feature; and determining whether the region of interest includes a high risk lesion.

Abstract: In one aspect, a method for inspecting features of an image using an image inspection controller that includes a processor communicatively coupled to a memory is described. The method includes receiving, at the processor, an input image, performing, on the input image, one of a semantic segmentation process and an object classification process to generate an output image, and prompting a user to select between approving the displayed output image, and at least one of i) performing an additional semantic segmentation process on the displayed output image, and ii) performing an additional object classification process on the displayed output image.

Abstract: An object-specific keypoint separation apparatus includes: an inference execution unit configured to receive a captured image capturing an object as an input and use a pre-trained model that has been trained in order to output a plurality of first maps and a plurality of second maps generated from the input captured image to output the plurality of first maps and the plurality of second maps, the plurality of first maps storing a vector describing a connection relationship of a keypoint of the object only around the keypoint, and the plurality of second maps representing a heat map configured to have a peak at coordinates at which the keypoint of the object appears; a map correction unit configured to correct the plurality of second maps using the plurality of first maps and the plurality of second maps; an upsampling unit configured to upsample the plurality of first maps; and an object-specific keypoint separation unit configured to separate keypoints for each object based on the plurality of upsampled firs

Abstract: An information processing apparatus (2) includes: a reducing unit (211) that generates a second image (IMG2) by reducing a first image (MGI) in which a target object is included; a first extracting unit (212) that extracts, as a first key point, a key point (KP) of the target object from the second image; a setting unit (213) that sets a target area (TA) that designates a part of the first image based on the first key point; and a second extracting unit (214) that extracts, as a second key point, a key point of the target object from a target image (IMG1 TA) of the first image that is included in the target area.