Abstract: An autonomous driving method based on prior knowledge of high-precision maps is provided. The autonomous driving method includes steps of: acquiring a current location of the autonomous driving vehicle; acquiring a prior-knowledge set associated with the current location from a high-precision map; acquiring sensing information by one or more sensing devices; acquiring one or more pieces of the prior-knowledge associated with the sensing information from the prior-knowledge set; calculating a control instruction according to one or more pieces of the prior-knowledge; controlling the autonomous driving vehicle driving according to the control command. Furthermore, an intelligent control device and an autonomous driving device are also provided.

Abstract: The invention provides a dynamic planning method of drivable area. The dynamic planning method of drivable area includes steps of: the first step of obtaining the current; the location of the autonomous driving vehicle at any time; sensing the week of the autonomous driving vehicle. An environment data of surrounding environment is extracted from the environment data. According to the current time location of the autonomous driving vehicle. The second step of acquiring autonomous driving vehicle is to set up high-definition map and lane information. Extracting information about static objects from environmental data. To extract information about dynamic objects from environmental data. Dynamic information is configured to predict the motion of dynamic objects. According to the first drivable area, static information and moving track. The second drivable area is planned based on the track and lane information.

Abstract: A sensing method for an autonomous driving vehicle includes steps of: obtaining an instruction; driving a center sensor assembly, or a center sensor assembly to sense corresponding areas to obtain first road condition data; judging whether a first occlusion area appears in the first road condition data; calculating a first safety area when a first occlusion area appears in the first road condition data; planning a first execution instruction; controlling the autonomous driving vehicle to motion in the first safety area; driving the at least one edge sensor assembly, or driving the at least one edge sensor assembly and the center sensor assembly to obtain second road condition data; judging whether a second occlusion area appears in the second road condition data; and planning a second execution instruction to control the autonomous driving vehicle to motion when a second occlusion area appears in the second road condition data.

Abstract: A simulation test method for autonomous driving vehicle includes steps of: obtaining source data; constructing a simulation scene according to the source data; controlling a virtual vehicle to perform a simulation test in the simulation scene; detecting a virtual driving trajectory of the virtual vehicle at a predetermined time interval while performing the simulation test; calculating difference between a first position and a second position; adjusting perceptual information related to a coordinate of the first position as perceptual information related to a coordinate of the second position to obtain modified perceptual information when the difference exceeds a first preset value; controlling the virtual vehicle to perform the simulation test based on the modified perceptual information, and detecting the virtual driving trajectory of the virtual vehicle at the predetermined time interval again. Furthermore, a computer equipment and a medium are also provided.

Abstract: An intelligent parking method of an autonomous driving vehicle is provided. The intelligent parking method includes steps of: recommending a first parking location according to an acquired target location; constructing an observation area according to the first parking location; obtaining environmental data of the observation area when the autonomous driving vehicle reaches the observation area; determining whether the first parking location is available or not according to the environmental data of the observation area; driving to the first parking location to park when the first parking location is available; or acquiring a second parking location according to a predetermined evaluation rule when the first parking location is unavailable, and recommending the one or more second parking locations to the user for selection; and driving to the second parking location selected by the user to park.

Abstract: A system and method for providing an autonomous delivery vehicle (ADV) incorporated with intelligent ramp control is disclosed. The ADV is configured to make decisions to deploy/retract the ramps depending on some conditions around the ADV. The ADV comprising a computing device including a means for executing artificial intelligence (AI) software, a ramp system comprising a plurality of ramps, and a sensor assembly in communication with the computing device to collect environmental data around the ADV. The data is communicated to the AI software, which is configured to analyze the environmental data to detect one or more obstacles proximate to the plurality of ramps of the ADV. Further, AI software determines a decision to deploy/retract at least one ramp based on the analysis and transmits the decision to the computing device. The computing device is configured to manipulate each ramp to deploy/retract based on the received decision.

Abstract: A system and method for navigating an autonomous driving vehicle (ADV) by capturing and analyzing information of a global scene and local objects around the ADV, is disclosed. The system comprises a sensor assembly incorporated on the ADV and a computing device in communication with the sensor assembly. The sensor assembly is configured to collect environmental data around the ADV. The computing device comprises a processor, and a memory unit for storing a predefined scene template and environmental data. The computing device is configured to process the environmental data to identify a moving and static object. The computing device is further configured to observe an environmental scene around the ADV. The observed environmental scene is aligned with a predefined scene template. Further, the predefined scene template is adjusted using the processed environmental data. The computing device provides instruction to control the vehicle based on the adjusted scene template.

Abstract: A system and method for autonomously transporting and delivering one or more commodities from drop-off point to recipient preferred environment, is disclosed. The system is configured to analyze an authorized person's voice command or request and executes the request. The system provides at least two modes of operation to deliver the commodities comprising mapped location delivery mode and human following delivery mode. The system is configured to analyze the command and extract the delivery location. The system is further configured to identify the location of the point of the interest based on object detection system and environment understanding system, to deliver the commodities. At human following delivery mode, the ADV follows the recipient and saves the location or path data once the preferred environment is located. At mapped location delivery mode, ADV maneuvers itself to the destination location by retrieving information on recipient's previous history of delivering commodities.

Abstract: A system and method for providing an autonomous delivery vehicle (ADV) incorporated with intelligent ramp control is disclosed. The ADV is configured to make decisions to deploy/retract the ramps depending on some conditions around the ADV. The ADV comprising a computing device including a means for executing artificial intelligence (AI) software, a ramp system comprising a plurality of ramps, and a sensor assembly in communication with the computing device to collect environmental data around the ADV. The data is communicated to the AI software, which is configured to analyze the environmental data to detect one or more obstacles proximate to the plurality of ramps of the ADV. Further, AI software determines a decision to deploy/retract at least one ramp based on the analysis and transmits the decision to the computing device. The computing device is configured to manipulate each ramp to deploy/retract based on the received decision.

Abstract: A system and method for autonomously transporting and delivering one or more commodities from drop-off point to recipient preferred environment, is disclosed. The system is configured to analyze an authorized person's voice command or request and executes the request. The system provides at least two modes of operation to deliver the commodities comprising mapped location delivery mode and human following delivery mode. The system is configured to analyze the command and extract the delivery location. The system is further configured to identify the location of the point of the interest based on object detection system and environment understanding system, to deliver the commodities. At human following delivery mode, the ADV follows the recipient and saves the location or path data once the preferred environment is located. At mapped location delivery mode, ADV maneuvers itself to the destination location by retrieving information on recipient's previous history of delivering commodities.

Abstract: A system and method for navigating an autonomous driving vehicle (ADV) by capturing and analyzing information of a global scene and local objects around the ADV, is disclosed. The system comprises a sensor assembly incorporated on the ADV and a computing device in communication with the sensor assembly. The sensor assembly is configured to collect environmental data around the ADV. The computing device comprises a processor, and a memory unit for storing a predefined scene template and environmental data. The computing device is configured to process the environmental data to identify a moving and static object. The computing device is further configured to observe an environmental scene around the ADV. The observed environmental scene is aligned with a predefined scene template. Further, the predefined scene template is adjusted using the processed environmental data. The computing device provides instruction to control the vehicle based on the adjusted scene template.

Abstract: A method for detecting an object in an image includes extracting a first feature vector from a first region of an image using a first subnetwork, determining a second region of the image by resizing the first region into a fixed ratio using a second subnetwork, wherein a size of the first region is smaller than a size of the second region, extracting a second feature vector from the second region of the image using the second subnetwork, classifying a class of the object using a third subnetwork on a basis of the first feature vector and the second feature vector, and determining the class of object in the first region according to a result of the classification, wherein the first subnetwork, the second subnetwork, and the third subnetwork form a neural network, wherein steps of the method are performed by a processor.

Abstract: The subject disclosure relates to generating three-dimensional façade models from images. In one aspect, a systematic decomposition schema is disclosed for a façade structured into a Direct Acyclic Graph and implemented as a top-down recursive subdivision and bottom-up merging. In a further aspect, optimization of façade depth operating in a façade surface and in the super-pixel level of a whole subdivision region is described.

Abstract: System and methods for preparing a model of an environment for display are provided. In some aspects, a system includes a transformation module configured to modify a three-dimensional (3D) model of an environment such that a target wall portion of the 3D model is more visible from a viewing direction after the modification than prior to the modification. The modification of the 3D model is based on constructive solid geometry and comprises at least one of i) a removal of at least one ceiling wall portion of the 3D model, ii) a thickening of at least one lateral wall portion of the 3D model, and iii) a removal of an obstructing wall portion. The obstructing wall portion blocks at least a partial view of the target wall portion from the viewing direction.

Abstract: System and methods for generating a model of an environment are provided. In some aspects, a system includes a layer module configured to identify one or more layers of the environment based on a plurality of three-dimensional (3D) points mapping the environment. The system also includes a layout module configured to generate a layout for each layer. Each layout includes a two-dimensional (2D) model of the environment. The system also includes a construction module configured to generate a 3D model of the environment based on the 2D model of each layout.

Abstract: The subject disclosure relates to generating models from images. In an aspect, multi-view semantic segmentation is provided to recognize and segment images at the pixel level into semantically meaningful areas, and which can provide labels with a specific object class. In further aspects, a partition scheme is provided that can separate objects into independent blocks using major line structures of a scene. In addition, an inverse patch-based orthographic composition and structure analysis on a block is provided that can regularize noisy and missing reconstructed 3D data to facilitate image-based modeling.

Abstract: The subject disclosure relates to generating models from images. In an aspect, multi-view semantic segmentation is provided to recognize and segment images at the pixel level into semantically meaningful areas, and which can provide labels with a specific object class. In further aspects, a partition scheme is provided that can separate objects into independent blocks using major line structures of a scene. In addition, an inverse patch-based orthographic composition and structure analysis on a block is provided that can regularize noisy and missing reconstructed 3D data to facilitate image-based modeling.

Abstract: The subject disclosure relates to generating three-dimensional façade models from images. In one aspect, a systematic decomposition schema is disclosed for a façade structured into a Direct Acyclic Graph and implemented as a top-down recursive subdivision and bottom-up merging. In a further aspect, optimization of façade depth operating in a façade surface and in the super-pixel level of a whole subdivision region is described.