Baidu USA LLC

Beijing Baidu Netcom Science Technology Co., Ltd

Apollo Intelligent Driving Technology (beijing) Co., Ltd.

Baidu.com Times Technology (beijing) Co. Ltd.

Jing-jin Electric Technologies Co., Ltd.

Capitalbio Corporation

Baidu Online Network Technology (beijing) Co., Ltd.

Beijing Baidu Netcom Science and Technology Co., Ltd.

Yao Zhou

Guowei Wan

Shiyu Song

Shenhua Hou

Liang Peng

Weixin Lu

Lei Wang

Hao Jia Li

Ping Yu

Wanli Xing

Fangfang Dong

Hu Wang

Dong Wang

Jing Cheng

Hao Li

Mingxian Lin

Yuzhe He

Xinliang Ling

Zhicheng Guo

Xiang Chen

Guoliang Huang

Liang Bai

Xiangyu Fu

Hantian Zhang

Tongjun Wang

A method for fusing road data to generate a map, includes: determining benchmark road data and at least one road data to be fused in a target road area; establishing, successively for each road data to be fused, a first road element association relationship between the first sub road data and the benchmark road data; establishing a second road element association relationship between the second sub road data and the benchmark road data according to the first road element association relationship; and fusing the benchmark road data and the road data to be fused according to the above association relationships to update the benchmark road data.

Method for fusing road data to generate a map, electronic device, and storage medium

A method of processing an image, a method of training a multi-task processing model, and an electronic device, which relate to a field of an automatic driving technology, in particular to a field of high-definition map technology. The method of processing an image includes: processing a to-be-processed image to obtain a feature point of the to-be-processed image, a feature point descriptor map of the to-be-processed image, and a dense descriptor map of the to-be-processed image; determining a pair of matched feature points between the to-be-processed image and a reference image based on the feature point and the feature point descriptor map; and determining a pair of matched pixels between the to-be-processed image and the reference image based on the dense descriptor map.

Method of processing image, method of training model, and electronic device

A positioning method based on a lane line and a feature point, an electronic device, and a storage medium, which relate to fields of computer, automatic driving, intelligent transportation, computer vision. The method may include: determining first real-time measurement residuals according to first sensor information of a movable object detected, the first real-time measurement residuals including a first inertial measurement unit measurement residual, a first lane line measurement residual and a first non-lane line measurement residual; updating a state vector of the movable object according to a kinematic model of an inertial measurement unit and the first real-time measurement residuals; and determining a pose of the movable object at a time instant corresponding to the updated state vector according to a pose vector in the updated state vector.

Positioning method based on lane line and feature point, electronic device, and storage medium

Exemplary positioning method and apparatus, a device, a system, a medium and a self-driving vehicle are provided. The positioning method includes determining, according to the current frame of point cloud data collected by a to-be-positioned terminal in a travelling environment, the current key point in the current frame of point cloud data and a point cloud distribution feature of the current key point; selecting, according to point cloud distribution features associated with reference key points in a global positioning map and the point cloud distribution feature of the current key point, a target key point matching the current key point from the reference key points; and determining, according to reference pose data associated with the target key point, the current pose data of the to-be-positioned terminal.

Positioning method and apparatus, device, system, medium and self-driving vehicle

The present disclosure provides a data processing method, an apparatus, an electronic device and a medium, which relates to the technical fields of autonomous driving, electronic maps, deep learning, image processing, and the like. The method includes: a computing device inputs a reference image and a captured image into a feature extraction model; obtain, a set of reference descriptors based on the first descriptor map; determine a plurality of sets of training descriptors; obtain a predicted pose of the vehicle by inputting the plurality of training poses and a plurality of similarities into a pose prediction model; and train the feature extraction model and the pose prediction model. When applied to a vehicle localization system, the trained feature extraction model and pose prediction model according to some embodiments of the present disclosure can improve accuracy and robustness of vehicle localization, thereby boosting the performance of the vehicle localization system.

Data processing method, data processing apparatus, electronic device and storage medium

A positioning method includes acquiring an image of an area where a target object is located at a first time instant and multiple frames of point cloud data of an area where the target object is located at multiple time instants, wherein the multiple frames of point cloud data include first point cloud data of the area where the target object is located at the first time instant. The method also includes determining a point cloud map according to the multiple frames of point cloud data and acquiring a target feature vector according to the first point cloud data, the point cloud map and the image. The method further includes determining a positioning result of the target object according to the target feature vector.

Positioning method, electronic device and computer readable storage medium

In one embodiment, a method for extracting point cloud features for use in localizing an autonomous driving vehicle (ADV) includes selecting a first set of keypoints from an online point cloud, the online point cloud generated by a LiDAR device on the ADV for a predicted pose of the ADV; and extracting a first set of feature descriptors from the first set of keypoints using a feature learning neural network running on the ADV, The method further includes locating a second set of keypoints on a pre-built point cloud map, each keypoint of the second set of keypoints corresponding to a keypoint of the first set of keypoint; extracting a second set of feature descriptors from the pre-built point cloud map; and estimating a position and orientation of the ADV based on the first set of feature descriptors, the second set of feature descriptors, and a predicted pose of the ADV.

Deep learning-based feature extraction for LiDAR localization of autonomous driving vehicles

An engine-and-electric-machine assembly includes an engine and an electric machine, a crankshaft being provided in the engine, the crankshaft including a main body and an extension section that extends out to the exterior of the engine, the extension section forming a rotation shaft of the electric machine, a rotor of the electric machine being mounted on the extension section, and a transition section being provided between the main body of the crankshaft and the extension section, wherein the rotor of the electric machine is connected to the transition section via a flange structure. By connecting the rotor of the electric machine and the crankshaft of the engine by using a flange, instead of using a key connecting structure, the strength of the connection between the rotor and the crankshaft can be improved, and optimize the moment of inertia of the transmission structure between the rotor and the crankshaft.

Engine-and-electric-machine assembly

In one embodiment, a method for solution inference using neural networks in LiDAR localization includes constructing a cost volume in a solution space for a predicted pose of an autonomous driving vehicle (ADV), the cost volume including a number of sub volumes, each sub volume representing a matching cost between a keypoint from an online point cloud and a corresponding keypoint on a pre-built point cloud map. The method further includes regularizing the cost volume using convention neural networks (CNNs) to refine the matching costs; and inferring, from the regularized cost volume, an optimal offset of the predicted pose. The optimal offset can be used to determine a location of the ADV.

LiDAR localization using 3D CNN network for solution inference in autonomous driving vehicles

Embodiments of the present disclosure disclose a method and apparatus for positioning an autonomous vehicle. The method includes: matching a current point cloud projected image of a first resolution with a map of the first resolution to generate a first histogram filter based on the matching result; determining at least two first response areas in the first histogram filter based on a probability value of an element in the first histogram filter; generating a second histogram filter based on a result of matching a current point cloud projected image of a second resolution with a map of the second resolution and the at least two first response areas, the first resolution being less than the second resolution; and calculating a weighted average of probability values of target elements in the second histogram filter to determine a positioning result of the autonomous vehicle in the map of the second resolution.

Method and apparatus for positioning autonomous vehicle

An integrated microfluidic chip, wherein at least one integrated reaction unit is provided on its substrate, and the integrated reaction unit comprises at least a sample cell (), a mixing cell () and a reaction cell () connected through liquid channels (). In one aspect, one end of the sample cell () is provided with a sample inlet (), and the chip further comprises an internal air circulating system/circuit. One end of the internal air circulating system/circuit is connected with the mixing cell (), while the other end comprises at least a first circulation branch circuit connected with the end of the sample cell () distal to the sample inlet ().

Integrated microfluidic chip and methods of use

In one embodiment, a method for temporal smoothness in localization results for an autonomous driving vehicle includes: creating a probability offset volume that represents an overall matching cost between a first set of keypoints from the online point cloud and a second set of keypoints from a pre-built point cloud map for each of a series of sequential light detection and ranging (LiDAR) frames in an online point cloud. The method also includes compressing the probability offset volume into multiple probability vectors across a X dimension, a Y dimension and a yaw dimension; providing each probability vector of the probability offset volume to a number of recurrent neural networks (RNNs); and generating, by the RNNs, a trajectory of location results across the plurality of sequential LiDAR frames.

Lidar localization using RNN and LSTM for temporal smoothness in autonomous driving vehicles

An engine-and-electric-machine assembly is provided that includes an engine and an electric machine, a crankshaft being provided in the engine, the crankshaft including a main body and an extension section that extends out to the exterior of the engine, the extension section forming a rotation shaft of the electric machine, and a rotor of the electric machine being mounted on the extension section. Moreover, a terminal of the rotation shaft is connected to a coolant pump, a rotor of the coolant pump is mounted to the rotation shaft, and while the rotation shaft is rotating the rotation shaft drives the coolant pump to provide coolant to the electric machine. By connecting the rotation shaft of the electric machine to the coolant pump, the assembly enables the pump to be highly integrated into the system and reduce manufacturing cost.

A microfluidic valve provided herein is configured to mix or capable of mixing a sample and/or a reagent in addition to controlling liquid flow. In one embodiment, the microfluidic valve comprises a rotor () and one or more micro-structures () that move with the rotation of the rotor (). In one embodiment, the one or more micro-structures () stir and/or mix content in a mixing chamber () formed by the rotor (), a base (), and a sleeve () of the microfluidic valve. A microfluidic chip or chip system comprising one or more of the microfluidic valves, and methods of use, are also provided.

Microfluidic valve and a chip or system comprising the microfluidic valve

The present disclosure provides a laser point cloud positioning method and system. The method comprises: converting laser point cloud reflection value data and height value data matched with a current location of an autonomous vehicle into laser point cloud projection data in a ground plane; assigning a weight for a reflection value matching probability and a height value matching probability of the laser point cloud projection data and a laser point cloud two-dimensional grid map, and determining a matching probability of the laser point cloud projection data and the laser point cloud two-dimensional grid map; determining a location of the autonomous vehicle in the laser point cloud two-dimensional grid map based on a matching probability of the laser point cloud projection data and the laser point cloud two-dimensional grid map. The present disclosure is employed to solve the problem in the prior art that when the laser point cloud matches with the map, an undesirable matching effect is achieved by individually considering the reflection value matching or height value matching, or superimposing the two simply. The present disclosure can improve the laser point cloud positioning precision, and enhance robustness of the positioning system.

Growing community of inventors

Beijing, China

Yao Zhou