Nec Corporation

Nec C Orporation

Quoc-Huy Tran

Manmohan Chandraker

Muhammad Zeeshan Zia

Pan Ji

Xiang Yu

Paul Vernaza

Bingbing Zhuang

Chi Li

Hyo Jin Kim

Mohammed E F Salem

Yi-Hsuan Tsai

Samuel Schulter

Buyu Liu

Mohammed E Fathy Salem

Yuliang Zou

Lokender Tiwari

A computer-implemented method executed by at least one processor for applying rolling shutter (RS)-aware spatially varying differential homography fields for simultaneous RS distortion removal and image stitching is presented. The method includes inputting two consecutive frames including RS distortions from a video stream, performing keypoint detection and matching to extract correspondences between the two consecutive frames, feeding the correspondences between the two consecutive frames into an RS-aware differential homography estimation component to filter out outlier correspondences, sending inlier correspondences to an RS-aware spatially varying differential homography field estimation component to compute an RS-aware spatially varying differential homography field, and using the RS-aware spatially varying differential homography field in an RS stitching and correction component to produce stitched images with removal of the RS distortions.

Joint rolling shutter image stitching and rectification

A method for jointly removing rolling shutter (RS) distortions and blur artifacts in a single input RS and blurred image is presented. The method includes generating a plurality of RS blurred images from a camera, synthesizing RS blurred images from a set of GS sharp images, corresponding GS sharp depth maps, and synthesized RS camera motions by employing a structure-and-motion-aware RS distortion and blur rendering module to generate training data to train a single-view joint RS correction and deblurring convolutional neural network (CNN), and predicting an RS rectified and deblurred image from the single input RS and blurred image by employing the single-view joint RS correction and deblurring CNN.

Joint rolling shutter correction and image deblurring

A method for improving geometry-based monocular structure from motion (SfM) by exploiting depth maps predicted by convolutional neural networks (CNNs) is presented. The method includes capturing a sequence of RGB images from an unlabeled monocular video stream obtained by a monocular camera, feeding the RGB images into a depth estimation/refinement module, outputting depth maps, feeding the depth maps and the RGB images to a pose estimation/refinement module, the depths maps and the RGB images collectively defining pseudo RGB-D images, outputting camera poses and point clouds, and constructing a 3D map of a surrounding environment displayed on a visualization device.

Pseudo RGB-D for self-improving monocular slam and depth prediction

A method is provided in an Advanced Driver-Assistance System (ADAS). The method extracts, from an input video stream including a plurality of images using a multi-task Convolutional Neural Network (CNN), shared features across different perception tasks. The perception tasks include object detection and other perception tasks. The method concurrently solves, using the multi-task CNN, the different perception tasks in a single pass by concurrently processing corresponding ones of the shared features by respective different branches of the multi-task CNN to provide a plurality of different perception task outputs. Each respective different branch corresponds to a respective one of the different perception tasks. The method forms a parametric representation of a driving scene as at least one top-view map responsive to the plurality of different perception task outputs. The method controls an operation of the vehicle for collision avoidance responsive to the at least one top-view map indicating an impending collision.

Multi-task perception network with applications to scene understanding and advanced driver-assistance system

A computer-implemented method for implementing a self-supervised visual odometry framework using long-term modeling includes, within a pose network of the self-supervised visual odometry framework including a plurality of pose encoders, a convolution long short-term memory (ConvLSTM) module having a first-layer ConvLSTM and a second-layer ConvLSTM, and a pose prediction layer, performing a first stage of training over a first image sequence using photometric loss, depth smoothness loss and pose cycle consistency loss, and performing a second stage of training to finetune the second-layer ConvLSTM over a second image sequence longer than the first image sequence.

Self-supervised visual odometry framework using long-term modeling and incremental learning

A method for correcting blur effects is presented. The method includes generating a plurality of images from a camera, synthesizing blurred images from sharp image counterparts to generate training data to train a structure-and-motion-aware convolutional neural network (CNN), and predicting a camera motion and a depth map from a single blurred image by employing the structure-and-motion-aware CNN to remove blurring from the single blurred image.

Image/video deblurring using convolutional neural networks with applications to SFM/SLAM with blurred images/videos

A method for correcting rolling shutter (RS) effects is presented. The method includes generating a plurality of images from a camera, synthesizing RS images from global shutter (GS) counterparts to generate training data to train the structure-and-motion-aware convolutional neural network (CNN), and predicting an RS camera motion and an RS depth map from a single RS image by employing a structure-and-motion-aware CNN to remove RS distortions from the single RS image.

Rolling shutter rectification in images/videos using convolutional neural networks with applications to SFM/SLAM with rolling shutter images/videos

A computer-implemented method, system, and computer program product are provided for a stabilization system utilizing pose estimation in an aerial drone. The method includes receiving, by a pose estimation system, a plurality of images from one or more cameras. The method also includes predicting, by the pose estimation system, a pose from the score map and a combined feature map, the combined feature map correlated from a pair of the plurality of images. The method additionally includes moving, by a propulsion system, the aerial drone responsive to the pose.

Aerial drone utilizing pose estimation

A computer-implemented method, system, and computer program product are provided for pose estimation. The method includes receiving, by a processor, a plurality of images from one or more cameras. The method also includes generating, by the processor with a feature extraction convolutional neural network (CNN), a feature map for each of the plurality of images. The method additionally includes estimating, by the processor with a feature weighting network, a score map from a pair of the feature maps. The method further includes predicting, by the processor with a pose estimation CNN, a pose from the score map and a combined feature map. The method also includes controlling an operation of a processor-based machine to change a state of the processor-based machine, responsive to the pose.

Learning good features for visual odometry

A method for estimating dense 3D geometric correspondences between two input point clouds by employing a 3D convolutional neural network (CNN) architecture is presented. The method includes, during a training phase, transforming the two input point clouds into truncated distance function voxel grid representations, feeding the truncated distance function voxel grid representations into individual feature extraction layers with tied weights, extracting low-level features from a first feature extraction layer, extracting high-level features from a second feature extraction layer, normalizing the extracted low-level features and high-level features, and applying deep supervision of multiple contrastive losses and multiple hard negative mining modules at the first and second feature extraction layers. The method further includes, during a testing phase, employing the high-level features capturing high-level semantic information to obtain coarse matching locations, and refining the coarse matching locations with the low-level features to capture low-level geometric information for estimating precise matching locations.

Dense three-dimensional correspondence estimation with multi-level metric learning and hierarchical matching

Systems and methods for detecting traffic scenarios include an image capturing device which captures two or more images of an area of a traffic environment with each image having a different view of vehicles and a road in the traffic environment. A hierarchical feature extractor concurrently extracts features at multiple neural network layers from each of the images, with the features including geometric features and semantic features, and for estimating correspondences between semantic features for each of the images and refining the estimated correspondences with correspondences between the geometric features of each of the images to generate refined correspondence estimates. A traffic localization module uses the refined correspondence estimates to determine locations of vehicles in the environment in three dimensions to automatically determine a traffic scenario according to the locations of vehicles. A notification device generates a notification of the traffic scenario.

Computer aided traffic enforcement using dense correspondence estimation with multi-level metric learning and hierarchical matching

Systems and methods for correspondence estimation and flexible ground modeling include communicating two-dimensional (2D) images of an environment to a correspondence estimation module, including a first image and a second image captured by an image capturing device. First features, including geometric features and semantic features, are hierarchically extract from the first image with a first convolutional neural network (CNN) according to activation map weights, and second features, including geometric features and semantic features, are hierarchically extracted from the second image with a second CNN according to the activation map weights. Correspondences between the first features and the second features are estimated, including hierarchical fusing of geometric correspondences and semantic correspondences. A 3-dimensional (3D) model of a terrain is estimated using the estimated correspondences belonging to the terrain surface. Relative locations of elements and objects in the environment are determined according to the 3D model of the terrain. A user is notified of the relative locations.

Dense correspondence estimation with multi-level metric learning and hierarchical matching

An action recognition system and method are provided. The action recognition system includes an image capture device configured to capture an actual image depicting an object. The action recognition system includes a processor configured to render, based on a set of 3D CAD models, synthetic images with corresponding intermediate shape concept labels. The processor is configured to form a multi-layer CNN which jointly models multiple intermediate shape concepts, based on the rendered synthetic images. The processor is configured to perform an intra-class appearance variation-aware and occlusion-aware 3D object parsing on the actual image by applying the CNN thereto to generate an image pair including a 2D and 3D geometric structure of the object. The processor is configured to control a device to perform a response action in response to an identification of an action performed by the object, wherein the identification of the action is based on the image pair.

Action recognition system with landmark localization on objects in images using convolutional neural networks

A system and method are provided for driving assistance. The system includes an image capture device configured to capture an actual image relative to an outward view from a motor vehicle and depicting an object. The system further includes a processor configured to render, based on a set of 3D CAD models, synthetic images with corresponding intermediate shape concept labels. The processor is further configured to form a multi-layer CNN which jointly models multiple intermediate shape concepts, based on the rendered synthetic images. The processor is also configured to perform an intra-class appearance variation-aware and occlusion-aware 3D object parsing on the actual image by applying the CNN to the actual image to output an image pair including a 2D and 3D geometric structure of the object. The processor is additionally configured to perform an action to mitigate a likelihood of harm involving the motor vehicle, based on the image pair.

Advanced driver-assistance system with landmark localization on objects in images using convolutional neural networks

A system and method are provided. The system includes an image capture device configured to capture an actual image depicting an object. The system also includes a processor. The processor is configured to render, based on a set of 3D Computer Aided Design (CAD) models, a set of synthetic images with corresponding intermediate shape concept labels. The processor is also configured to form a multi-layer Convolutional Neural Network (CNN) which jointly models multiple intermediate shape concepts, based on the rendered synthetic images. The processor is further configured to perform an intra-class appearance variation-aware and occlusion-aware 3D object parsing on the actual image by applying the CNN to the actual image to output an image pair including a 2D geometric structure and a 3D geometric structure of the object depicted in the actual image.

Growing community of inventors

Santa Clara, CA, United States of America

Quoc-Huy Tran