Mitsubishi Electric Research Laboratories, Inc.

Other

Hewlett-packard Development Company, L.p.

Columbia University

Mitsubishi Denki Kabushiki Kaisha

Michael J Jones

Tim Marks

Paul A Viola

Daniel Nikolaev Nikovski

Oncel C Tuzel

Jay E Thornton

Andrei Kniazev

Pieter J Van Zee

Manlik Kwong

Bharat Singh

Kathryn L Mitchell

Brian E Cripe

Gabe Beged-Dov

Ejaz Ahmed

Thomas C McFarland

Derek Roller

Rohith Mv

Raphael A Pelossof

James R Miller

Gabriel Beged-Dov

Kathy Shelby

Suren Kumar

Ritwik Kumar

Chavdar Papazov

Yanfeng Geng

Guodong Guo

Weiguang Si

Russell Mull

Alexander Broad

Kuldeep Kulkarni

Thomas Mcfarland

An object detector includes an input interface to accept a sequence of video frames, a memory to store a neural network trained to detect objects in the video frames, a processor to process each video frame sequentially with the neural network to detect objects in the sequence of video frames, and an output interface to output the object detection information. The neural network includes a first subnetwork, a second subnetwork, and a third subnetwork. The first subnetwork receives as an input a video frame and outputs a feature map of the video frame. The second subnetwork is a recurrent neural network that takes the feature map as an input and outputs a temporal feature map. The third subnetwork takes the temporal feature map as an input and outputs object detection information.

System and method for detecting objects in video sequences

A system for video anomaly detection partitions the input video into a set of input spatio-temporal regions according to parameters of the spatio-temporal regions of the training video indicative of a number of regions in each video frame defining a spatial dimension of each of the spatio-temporal regions and a number of video frames defining a temporal dimension of each of the spatio-temporal regions, and determines blurred, thresholded difference images for each of the input spatio-temporal regions to produce a set of blurred, thresholded difference images. Next, the system extracts a feature vector from each set of blurred, thresholded difference images to produce a set of input feature vectors, computes a smallest distance between each input feature vector and the training feature vectors corresponding to the same spatial region in the scene to produce a set of distances, and compares each distance from the set of distances with an anomaly detection threshold to detect anomalies in the input video of the scene.

System and method for detecting motion anomalies in video

A system for video anomaly detection includes an input interface configured to accept an input video of a scene, and a memory configured to store training video patches of a training video of the scene capturing normal activity in the scene, and store a neural network trained to compare two video patches to declare the compared video patches as similar or dissimilar. The system also includes a processor configured to partition the input video into input video patches, compare, using the neural network, each input video patch with corresponding training video patches retrieved from the memory to determine if each input video is similar to at least one corresponding training video patch, and declare an anomaly when at least one input video patch is dissimilar to all corresponding training video patches.

System and method for detecting anomalies in video using a similarity function trained by machine learning

A method and system determine a three-dimensional (3D) pose of an object and 3D locations of landmark points of the object by first obtaining a 3D point cloud of the object. 3D surface patches are extracted from the 3D point cloud, and a parametric model is fitted to each 3D surface patch to determine a set of descriptors. A set of correspondences between the set of descriptors and a set of descriptors of patches extracted from 3D point clouds of objects from the same object class with known 3D poses and known 3D locations of landmark points is determined. Then, the 3D pose of the object and 3D locations of the landmark points of the object are estimated from the set of correspondences.

Method and system for determining 3D object poses and landmark points using surface patches

An image processing system includes a memory to store data indicative of benchmark feature vectors of benchmark images, an input interface to accept data indicative of a first feature vector of a first image and a second feature vector of a second image, and an output interface to render a similarity value between the first and the second images. The system includes a processor to determine the similarity using a first hyperplane separating the benchmark feature vectors from the first feature vector and a second hyperplane separating the benchmark feature vectors from the second feature vector. The processor determines a first normal vector to the first hyperplane as the difference between the first feature vector and the mean of the benchmark feature vectors. The processor determines an offset for the first hyperplane as the average of the maximum inner product of the benchmark feature vectors with the first normal vector and the inner product of the first feature vector with the first normal vector. The processor determines the similarity value as a function of a sum of a signed distance of the second feature vector to the first hyperplane and a signed distance of the first feature vector to the second hyperplane.

System and method for image comparison based on hyperplanes similarity

A method and system detects actions of an object in a scene by first acquiring a video of the scene as a sequence of images, wherein each image includes pixels, wherein the video is partitioned into chunks. The object in the video is tracked. For each object and each chunk of the video, trajectories of the pixels within a bounding box located over the object are tracked, and cropped trajectories and cropped images for one or more images in the chunk are produced using the bounding box. Then, the cropped trajectories and cropped images are passed to a recurrent neural network (RNN) that outputs a relative score for each action of interest.

Method and system for detecting actions in videos

A method and system for detecting actions of an object in a scene from a video of the scene. The video is a video sequence partitioned into chunks, and each chunk includes consecutive video frames. The method including the following elements. Acquiring the video of the scene, wherein the video includes a sequence of images. Tracking the object in the video, and for each object and each chunk of the video, further comprising: determining a sequence of contour images from video frames of the video sequence to represent motion data within a bounding box located around the object. Using the bounding box to produce cropped contour images and cropped images for one or more images in each chunk. Passing the cropped contour images and the cropped images to a recurrent neural network (RNN) that outputs a relative score for each action of interest.

Method and system for detecting actions in videos using contour sequences

Anomalies in real time series are detected by first determining a similarity matrix of pairwise similarities between pairs of normal time series data. A spectral clustering procedure is applied to the similarity matrix to partition variables representing dimensions of the time series data into mutually exclusive groups. A model of normal behavior is estimated for each group. Then, for the real time series data, an anomaly score is determined, using the model for each group, and the anomaly score is compared to a predetermined threshold to signal the anomaly.

Method for anomaly detection in time series data based on spectral partitioning

A method detects anomalies in time series data, by first learning a final set of exemplars by summarizing training time series data using a divide-and-conquer procedure. Then, for each window of testing time series data, a distance to a nearest exemplar in the final set of exemplars is determined, wherein the distance is an anomaly score. Finally, an anomaly is signaled when the anomaly score for a window is greater than a threshold.

Method for learning exemplars for anomaly detection

Image capturing systems and methods of capturing images, especially in contact prone environments are provided. The image capturing systems includes an optical sensor and output device adapted to receive the electrical signals from optical sensor and transmit a corresponding signal, and a housing retaining the optical sensor and the output device. The housing is adapted to mount to a surface, for example, inside or on top of a helmet, and includes a low profile above the surface and a footprint in contact with the surface. The size of the footprint is selected whereby and impact loads on the housing are minimized or attenuated before being transmitted to the surface to which the housing is mounted. The invention is uniquely applicable to head gear mounting, but may be used on a broad range of article and a broad range of fields.

Image capturing system and method of use

A method re-identifies objects in a pair of images by applying a convolutional neural network (CNN). Each layer in the network operates on an output of a previous layer. The layers include a first convolutional layer and a first max pooling layer to determine a feature map, a cross-input neighborhood differences layer to produce neighborhood difference maps, a patch summary layer to produce patch summary feature maps, a first fully connected layer to produce a feature vector representing higher order relationships in the patch summary feature maps, a second fully connected layer to produce two scores representing positive pair and negative pair classes, and a softmax layer to produce positive pair and negative pair probabilities. Then, the positive pair probability is output to signal whether the two images represent the same object or not.

Method for determining similarity of objects represented in images

A method and system tracks objects in an environment by acquiring a first sequence of images of the environment with a visible-light camera having a first frame rate and a second sequence of images with a thermal infrared sensor having a second frame rate. The second frame rate is substantially lower than the first frame rate. The objects are tracked in the first sequence of images to obtain tracks. Warm regions in the second sequence of images are detected to obtain detections. The tracks and the detections are aligned spatially and temporally, and verified to determine whether the tracks and detections coincide after the aligning.

Method and system for tracking people in indoor environments using a visible light camera and a low-frame-rate infrared sensor

In a sequence of images of a scene acquired by a stationary camera, objects are detected and tracked by determining a first set of candidate foreground regions according to a background model. A second set of candidate foreground regions is determined according to a set of foreground models. Then, candidate foreground regions in the first set and the second set are validated to produce a final set of foreground regions in the image that include the objects.

Method for detecting and tracking objects in image sequences of scenes acquired by a stationary camera

A method detects anomalies in time series data by comparing universal features extracted from testing time series data with the universal features acquired from training time series data to determine a score. The universal features characterize trajectory components of the time series data and stochastic components of the time series data. Then, an anomaly is detected if the anomaly score is above a threshold.

Method for detecting anomalies in a time series data with trajectory and stochastic components

Future travel times of a target vehicle traveling on a route from a starting point to a destination are predicted by first acquiring, by a probe vehicle, real-time probe data to alternative links from the starting point to the destination. Then, the future travel time for each link is predicted using a set of regression functions.

Growing community of inventors

Cambridge, MA, United States of America

Michael J Jones