Abstract: A method of lane detection for a non-transitory computer readable storage medium storing one or more programs is disclosed. The one or more programs include instructions, which when executed by a computing device, cause the computing device to perform the following steps comprising: generating a ground truth; off-line training a lane detection algorithm by using the ground truth, the lane detection algorithm using parameters that express a lane marking in an arc; on-line generating a predicted lane marking; comparing the predicted lane marking against the ground truth; and off-line refining the lane detection algorithm.

Abstract: A system for generating a ground truth dataset for motion planning of a vehicle is disclosed. The system includes an internet server that further includes an I/O port, configured to transmit and receive electrical signals to and from a client device; a memory; one or more processing units; and one or more programs stored in the memory and configured for execution by the one or more processing units, the one or more programs including instructions for: a corresponding module configured to correspond, for each pair of images, a first image of the pair to a LiDAR static-scene point cloud; and a computing module configured to compute a camera pose associated with the pair of images in the coordinate of the point cloud.

Abstract: A system and method for implementing an image annotation platform are disclosed. A particular embodiment includes: registering a plurality of labelers to which annotation tasks are assigned; assigning annotation tasks to the plurality of labelers; determining if the annotation tasks can be closed or re-assigned to the plurality of labelers; aggregating annotations provided by the plurality of labelers as a result of the closed annotation tasks; evaluating a level of performance of the plurality of labelers in providing the annotations; and calculating payments for the plurality of labelers based on the quantity and quality of the annotations provided by the plurality of labelers.

Abstract: A Method and System for Modeling Autonomous Vehicle Behavior. The system and method makes it feasible to develop an autonomous vehicle control system for complex vehicles, such as for cargo trucks and other large payload vehicles. The method and system commences by first obtaining 3-dimensional data for one or more sections of roadway. Once the 3-dimensional roadway data is obtained, that data is used to run computer simulations of a computer model of a specific vehicle being controlled by a generic vehicle control algorithm or system. The generic vehicle control algorithm is optimized by running the simulations utilizing the 3-dimensional roadway data until an acceptable performance result is acheived. Once an acceptable simulation is executed using the generic vehicle control algorithm, the control algorithm/system is used to run one or more real-world driving tests on the roadway for which the 3-dimensional data was obtained.

Abstract: A system and method for vehicle occlusion detection is disclosed.

Abstract: A system and method for using human driving patterns to detect and correct abnormal driving behaviors of autonomous vehicles are disclosed. A particular embodiment includes: generating data corresponding to a normal driving behavior safe zone; receiving a proposed vehicle control command; comparing the proposed vehicle control command with the normal driving behavior safe zone; and issuing a warning alert if the proposed vehicle control command is outside of the normal driving behavior safe zone. Another embodiment includes modifying the proposed vehicle control command to produce a modified and validated vehicle control command if the proposed vehicle control command is outside of the normal driving behavior safe zone.

Abstract: A system and method for using triplet loss for proposal free instance-wise semantic segmentation for lane detection are disclosed. A particular embodiment includes: receiving image data from an image generating device mounted on an autonomous vehicle; performing a semantic segmentation operation or other object detection on the received image data to identify and label objects in the image data with object category labels on a per-pixel basis and producing corresponding semantic segmentation prediction data; performing a triplet loss calculation operation using the semantic segmentation prediction data to identify different instances of objects with similar object category labels found in the image data; and determining an appropriate vehicle control action for the autonomous vehicle based on the different instances of objects identified in the image data.

Abstract: A system and method for distributed graphics processing unit (GPU) computation are disclosed. A particular embodiment includes: receiving a user task service request from a user node; querying resource availability from a plurality of slave nodes having a plurality of graphics processing units (GPUs) thereon; assigning the user task service request to a plurality of available GPUs based on the resource availability and resource requirements of the user task service request, the assigning including starting a service on a GPU using a distributed processing container and creating a corresponding uniform resource locator (URL); and retaining a list of URLs corresponding to the resources assigned to the user task service request.

Abstract: A method of visual odometry for a non-transitory computer readable storage medium storing one or more programs is disclosed. The one or more programs include instructions, which when executed by a computing device, causes the computing device to perform the following steps comprising: in response to images in pairs, generating a prediction of static scene optical flow for each pair of the images in a visual odometry model; generating a set of motion parameters for each pair of the images in the visual odometry model; training the visual odometry model by using the prediction of static scene optical flow and the motion parameters; and predicting motion between a pair of consecutive image frames by the trained visual odometry model.

Abstract: A method of localization for a non-transitory computer readable storage medium storing one or more programs is disclosed. The one or more programs comprise instructions, which when executed by a computing device, cause the computing device to perform utilizing one or more autonomous vehicle driving modules that execute processing of images from a camera and data from a LiDAR the following steps comprising: aligning a 3D submap with a global map; extracting features from the 3D submap and the global map; classifying the extracted features in classes; and establishing correspondence of features in a same class between the 3D submap and the global map.

Abstract: A method of localization for a non-transitory computer readable storage medium storing one or more programs is disclosed. The one or more programs comprise instructions, which when executed by a computing device, cause the computing device to perform by one or more autonomous vehicle driving modules execution of processing of images from a camera and data from a LiDAR using the following steps comprising: constructing a 3D submap and a global map; extracting features from the 3D submap and the global map; matching features extracted from the 3D submap against features extracted from the global map; refining feature correspondence; and refining location of the 3D submap.

Abstract: A system and method for semantic segmentation using hybrid dilated convolution (HDC) are disclosed. A particular embodiment includes: receiving an input image; producing a feature map from the input image; performing a convolution operation on the feature map and producing multiple convolution layers; grouping the multiple convolution layers into a plurality of groups; applying different dilation rates for different convolution layers in a single group of the plurality of groups; and applying a same dilation rate setting across all groups of the plurality of groups.

Abstract: A system method for occluding contour detection using a fully convolutional neural network is disclosed. A particular embodiment includes: receiving an input image; producing a feature map from the input image by semantic segmentation; applying a Dense Upsampling Convolution (DUC) operation on the feature map to produce contour information of objects and object instances detected in the input image; and applying the contour information onto the input image.

Abstract: A system and method for traffic data collection using unmanned aerial vehicles (UAVs) are disclosed.

Abstract: A system and method for semantic segmentation using dense upsampling convolution (DUC) are disclosed. A particular embodiment includes: receiving an input image; producing a feature map from the input image; performing a convolution operation on the feature map and reshape the feature map to produce a label map; dividing the label map into equal subparts, which have the same height and width as the feature map; stacking the subparts of the label map to produce a whole label map; and applying a convolution operation directly between the feature map and the whole label map without inserting extra values in deconvolutional layers to produce a semantic label map.