Abstract: A method includes receiving an image comprising a human head from an imaging device. Further, the method includes grouping one or more pixels of the image into one or more super pixels. Furthermore, the method includes generating a polar coordinate grid for the image, wherein the polar coordinate grid comprises one or more grid points. Additionally, the method includes labeling the one or more grid points as one of a face region, a hair region, and a background region based on tiered structure constraint. The method also includes labeling the one or more super pixels as one of the face region, the hair region, and the background region based on the one or more labels of the one or more grid points. Further, the method includes transmitting a notification comprising the one or more labels of the one or more super pixels.

Abstract: A method for detecting and tracking a target includes detecting the target using a plurality of feature cues, fusing the plurality of feature cues to form a set of target hypotheses, tracking the target based on the set of target hypotheses and a scene context analysis, and updating the tracking of the target based on a target motion model.

Abstract: According to some embodiments, a signal processing unit may receive distributed acoustic sensing data associated with a first set of a plurality of pipeline locations. The signal processing unit may also receive collected physical data representing a physical characteristic of a second set of a plurality of pipeline locations. The signal processing unit may then utilize a pipeline model having the distributed acoustic sensing data and collected physical data as inputs to automatically generate at least one alert indicating an increased probability of damage to the pipeline.

Abstract: A method for analyzing time series data to identify an event of interest is provided. The method includes receiving, at a computing device, time series data that includes the event of interest, identifying, using the computing device, a start time of the event of interest and an end time of the event of interest by modeling at least one transitional pattern in the time series data, and categorizing, using the computing device, the event of interest based on the at least one transitional pattern.

Abstract: A novel technique for unsupervised feature selection is disclosed. The disclosed methods include automatically selecting a subset of a feature of an image. Additionally, the selection of the subset of features may be incorporated with a congealing algorithm, such as a least-square-based congealing algorithm. By selecting a subset of the feature representation of an image, redundant and/or irrelevant features may be reduced or removed, and the efficiency and accuracy of least-square-based congealing may be improved.

Abstract: Systems provided herein include a learning environment and an agent. The learning environment includes an avatar and an object. A state signal corresponding to a state of the learning environment includes a location and orientation of the avatar and the object. The agent is adapted to receive the state signal, to issue an action capable of generating at least one change in the state of the learning environment, to produce a set of observations relevant to a task, to hypothesize a set of action models configured to explain the observations, and to vet the set of action models to identify a learned model for the task.

Abstract: A novel technique for performing video matting, which is built upon a proposed image matting algorithm that is fully automatic is disclosed. The disclosed methods utilize a PCA-based shape model as a prior for guiding the matting process, so that manual interactions required by most existing image matting methods are unnecessary. By applying the image matting algorithm to these foreground windows, on a per frame basis, a fully automated video matting process is attainable. The process of aligning the shape model with the object is simultaneously optimized based on a quadratic cost function.

Abstract: A tracking and recognition system is provided. The system includes a computer vision-based identity recognition system configured to recognize one or more persons, without a priori knowledge of the respective persons, via an online discriminative learning of appearance signature models of the respective persons. The computer vision-based identity recognition system includes a memory physically encoding one or more routines, which when executed, cause the performance of constructing pairwise constraints between the unlabeled tracking samples. The computer vision-based identity recognition system also includes a processor configured to receive unlabeled tracking samples collected from one or more person trackers and to execute the routines stored in the memory via one or more algorithms to construct the pairwise constraints between the unlabeled tracking samples.

Abstract: A novel technique for performing video matting, which is built upon a proposed image matting algorithm that is fully automatic is disclosed. The disclosed methods utilize a PCA-based shape model as a prior for guiding the matting process, so that manual interactions required by most existing image matting methods are unnecessary. By applying the image matting algorithm to these foreground windows, on a per frame basis, a fully automated video matting process is attainable. The process of aligning the shape model with the object is simultaneously optimized based on a quadratic cost function.

Abstract: In accordance with one aspect of the present technique, a method is disclosed. The method includes receiving a new video from one or more sensors and generating a new content graph (CG) based on the new video. The method also includes comparing the new CG with a plurality of prior CGs. The method further includes identifying a first portion of the new CG matching a portion of a first prior CG and a second portion of the new CG matching a portion of the second prior CG. The method further includes analyzing a first set of semantic annotations (SAs) associated with the portion of the first prior CG and a second set of SAs associated with the portion of the second prior CG. The method further includes generating a sequence of SAs for the new video based on the analysis of the first and the second set of SAs.

Abstract: An example method includes: classifying lung function risk based on patient attributes and a clinical protocol; generating alarms and incentives for compliance with the clinical protocol based on patient attributes, clinical protocol, and patient lung function risk; determining an orientation and position of a clinical device based on tagged feature(s) of the clinical device compared to identified patient feature(s); monitoring patient interaction with the clinical device; identifying a deviation from the clinical protocol based on the monitored patient interaction, a patient biometric indicator, and a desired setpoint state in the protocol; when a deviation is identified, providing feedback proportional to the deviation, the feedback including an adjustment with respect to the clinical protocol and/or the clinical device; and triggering at least one alarm and/or incentive based on deviation and feedback, wherein the alarm/incentives differs based on whether and to what extent deviation is identified and feed

Abstract: A novel technique for unsupervised feature selection is disclosed. The disclosed methods include automatically selecting a subset of a feature of an image. Additionally, the selection of the subset of features may be incorporated with a congealing algorithm, such as a least-square-based congealing algorithm. By selecting a subset of the feature representation of an image, redundant and/or irrelevant features may be reduced or removed, and the efficiency and accuracy of least-square-based congealing may be improved.

Abstract: A novel technique for unsupervised feature selection is disclosed. The disclosed methods include automatically selecting a subset of a feature of an image. Additionally, the selection of the subset of features may be incorporated with a congealing algorithm, such as a least-square-based congealing algorithm. By selecting a subset of the feature representation of an image, redundant and/or irrelevant features may be reduced or removed, and the efficiency and accuracy of least-square-based congealing may be improved.

Abstract: A method for identifying assets is provided. The method includes illuminating an identification pattern associated with the asset with non-visible light. The identification pattern includes retroreflective material that represents identity information of the asset. The method further includes adjusting either a shutter speed of an image sensor, or a power of the non-visible light, or both, based on motion state of the asset. The image sensor is configured to receive response of the retroreflective material to the non-visible light. Furthermore, the method includes the step of identifying the asset based on decoded identity information. The decoded identity information is obtained by processing the response of the retroreflective material.

Abstract: A system and method for estimating a set of landmarks for a large image ensemble employs only a small number of manually labeled images from the ensemble and avoids labor-intensive and error-prone object detection, tracking and alignment learning task limitations associated with manual image labeling techniques. A semi-supervised least squares congealing approach is employed to minimize an objective function defined on both labeled and unlabeled images. A shape model is learned on-line to constrain the landmark configuration. A partitioning strategy allows coarse-to-fine landmark estimation.

Abstract: A novel technique for unsupervised feature selection is disclosed. The disclosed methods include automatically selecting a subset of a feature of an image. Additionally, the selection of the subset of features may be incorporated with a congealing algorithm, such as a least-square-based congealing algorithm. By selecting a subset of the feature representation of an image, redundant and/or irrelevant features may be reduced or removed, and the efficiency and accuracy of least-square-based congealing may be improved.

Abstract: A method for detecting and tracking a target includes detecting the target using a plurality of feature cues, fusing the plurality of feature cues to form a set of target hypotheses, tracking the target based on the set of target hypotheses and a scene context analysis, and updating the tracking of the target based on a target motion model.

Abstract: A novel technique for unsupervised feature selection is disclosed. The disclosed methods include automatically selecting a subset of a feature of an image. Additionally, the selection of the subset of features may be incorporated with a congealing algorithm, such as a least-square-based congealing algorithm. By selecting a subset of the feature representation of an image, redundant and/or irrelevant features may be reduced or removed, and the efficiency and accuracy of least-square-based congealing may be improved.

Abstract: A method for image alignment is disclosed. In one embodiment, the method includes acquiring a facial image of a person and using a discriminative face alignment model to fit a generic facial mesh to the facial image to facilitate locating of facial features. The discriminative face alignment model may include a generative shape model component and a discriminative appearance model component. Further, the discriminative appearance model component may have been trained to estimate a score function that minimizes the angle between a gradient direction and a vector pointing toward a ground-truth shape parameter. Additional methods, systems, and articles of manufacture are also disclosed.

Abstract: Homography-based imaging apparatus and method are provided. The apparatus may include a processor (44) coupled to process respective sequences of sky images respectively acquired by physical image acquisition devices 181 and 182 at respective spaced apart locations (e.g., P1, P2). The processor may include an image alignment module (32) configured to spatially relate respective views of at least one object (e.g., clouds, aerial vehicles) visible in the respective sequences of the sky images based on homography (42) of at least one astronomical image acquired at each spaced apart location. The astronomical image may include a number of spatial references corresponding to respective astronomical body positions located practically at infinity relative to a respective distance between the spaced apart locations. Further views (synthetic views) may be generated at selectable new locations (e.g., P3, P4, P5, P6), without actually having any physical image acquisition devices at such selectable locations.