Abstract: A unified learning approach for large-scale ride-hailing is described, the approach includes obtaining an offline state value network for predicting a value of a vehicle state, the offline state value network being trained based on a plurality of historical vehicle trajectories; initializing an online state value network and dispatching a plurality of vehicles according to the online state value network for a period of time; training the online state value network based on vehicle states of the plurality of vehicles before and after the dispatching and rewards associated with the dispatching; ensembling the trained online state value network and the offline state value network to obtain an ensembled online state value network; and dispatching the plurality of vehicles according to the ensembled online state value network.

Abstract: The present disclosure is related to systems and methods for indoor positioning. The method includes obtaining a first location of a subject at a first time point. The method also includes determining one or more motion parameters associated with the subject at least based on one or more sensors carried by the subject. The method further includes determining, based on the first location of the subject and the one or more motion parameters associated with the subject, a second location of the subject at a second time point after the first time point. The method still further includes adjusting, at least based on surroundings information associated with the subject at the second time point, the second location of the subject to determine a target location of the subject.

Abstract: Systems and methods for generating a road map are provided. The methods include obtaining a navigation route for a navigation terminal and a current location of the navigation terminal. The methods further include obtaining road network data associated with the current location of the navigation terminal based on the navigation route and the current location of the navigation terminal. The road network data includes a plurality of links and a plurality of nodes. Each node is connected to one or more links including at least one type of links. The methods further include determining a level for each node. The methods further include determining one or more target links for the at least one type of links connected to the node based on the level of the node. The methods further include generating a road map based on the target links.

Abstract: The application discloses Systems and methods for a data synchronization. The system may include a receiving module, an instruction generating module and a sending module. The receiving module may be configured to receive the first instruction. The first instruction may be used to instruct the start of data acquisition of the system. In response to receiving the first instruction, the instruction generating module may be configured to generate a second instruction. The second instruction may be used to trigger at least two sensors to acquire data. The sending module may be configured to send second instruction to at least two sensors respectively based on the first delay. The first delay causes the time difference between at least two sensors starting to acquire data less than the first preset threshold.

Abstract: The present disclosure relates to an information providing method for an on-demand service. The method may include receiving service request information from a passenger of a passenger terminal device. The service request information may include a departure location of the passenger. The method may further include acquiring historical service request information related to the passenger; and determining travel-route-related information based at least in part on the departure location of the passenger and the historical service request information. Also disclosed is a system for implementing the method.

Abstract: The present disclosure is related to systems and methods for processing speech dialogue. The method includes obtaining target speech dialogue data. The method includes obtaining a text vector representation sequence, a phonetic symbol vector representation sequence, and a role vector representation sequence by performing a vector transformation on the target speech dialogue data based on a text embedding model, a phonetic symbol embedding model, and a role embedding model, respectively. The method includes determining a representation vector corresponding to the target speech dialogue data by inputting the text vector representation sequence, the phonetic symbol vector representation sequence, and the role vector representation sequence into a trained speech dialogue coding model. The method includes determining a summary of the target speech dialogue data by inputting the representation vector into a classification model.

Abstract: A system and method of determining a policy to prevent fading drivers is described. The system and method creates virtual trajectories of incentives such as coupons offered to drivers in a transportation hailing system and corresponding states of drivers in response to the incentives. A joint policy simulator is created from an incentive policy, a confounding incentive policy, and an incentive object policy to generate the simulated actions of drivers in response to different incentives. The rewards of the simulated actions of the drivers is determined by a discriminator. The incentive policy for preventing fading drivers is optimized by reinforcement learning based on the virtual trajectories generated by the joint policy simulator and discriminator.

Abstract: After determining a trip order has a high risk, verification information unavailable in a passenger information database system can be requested and provided for a passenger of the account. Validity of the verification information can be determined using a verification service and stored in a passenger information database. A decision for the trip order can be determined based on the validity of the verification information stored in the passenger information database.

Abstract: Systems and methods for updating a model of a building are provided. The method includes obtaining a three-dimensional (3D) model of a building including reference environmental data associated with a plurality of reference locations of the building and obtaining current environmental data corresponding to a current location of the building. The method further includes identifying, from the 3D model, a target reference location corresponding to the current location from the plurality of reference locations of the building and comparing the current environmental data with the target reference environmental data. The method further includes sending to the user device the target reference location when the current environmental data match the target reference environmental data, and updating the 3D model with the current environmental data when the current environmental data statistically un-match the target reference environmental data.

Abstract: Embodiments of the disclosure provide systems and methods for performing motion transfer using a learning model. An exemplary system may include a communication interface configured to receive a first image including a first movable object and a second image including a second movable object. The system may also include at least one processor coupled to the communication interface. The at least one processor may be configured to extract a first set of motion features of the first movable object from the first image using a first encoder of the learning model and extract a first set of static features of the second movable object from the second image using a second encoder of the learning model. The at least one processor may also be configured to generate a third image by synthesizing the first set of motion features and the first set of static features.

Abstract: Incidents may be monitored using a camera in response to an activation of a trigger device. Trigger data may be received over a communication network in response to activation of a trigger device disposed within a first area of a vehicle. A camera disposed within a second area of the vehicle may be controlled over the communication network in response to receiving the trigger data, the controlling causing the camera to capture sensor information of a passenger of the vehicle. Passenger information of the passenger of the vehicle may be obtained in response to receiving the trigger data. An alert notification may be generated. The alert notification may comprise at least a portion of the sensor information of the passenger, and the alert notification may be capable of initiating one or more incident actions associated with the passenger.

Abstract: The present disclosure relates to a system and a method for electric vehicle charging. The system may perform the method to: receive a request for charging station recommendation from a client terminal; determine at least one charging station within a preset distance range of the client terminal; determine an evaluation for each of the at least one charging station based on one or more evaluation parameters; and transmitting, by the first computing device, the evaluation of the each of the at least one charging station and an identifier of the each of the at least one charging station to the client terminal.

Abstract: The present disclosure provides a method for pose determination. The method may include obtaining a first pose of a subject at a first time point. The method may further include retrieving one or more first features associated with a road from a database. The road may be viewable by the subject at the first pose and at the first time point. The method may further include obtaining an image captured by the subject at a second time point. The method may further include extracting one or more second features associated with the road from the image. The method may also include determining a second pose of the subject at the second time point at least based on comparing the one or more first features associated with the road with the one or more second features associated with the road.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for a computer-implemented method for displaying a heat map are provided. One of the methods includes: receiving heat data from a server, wherein the heat data comprises a color value associated with each of a plurality of regions and one or more geographic locations respectively corresponding to one or more boundary points of each of one or more heat regions among the plurality of regions; generating a first heat map comprising a plurality of pixels; obtaining a second heat map by adjusting the color value of each pixel in the first heat map; and providing the second heat map for display, wherein the second heat is superimposed on a digital map. This method addresses the issue of conventional techniques that service request information may be misrepresented due to abrupt changes in color values across neighboring heat regions.

Abstract: The present disclosure relates to systems and methods for identifying one or more target roads. The systems may perform the methods to obtain a heat map associated with a plurality of driving track points along a plurality of roads in a target region, wherein the plurality of roads includes one or more target roads and one or more reference roads; obtain a road network map associated with the one or more reference roads in the target region; produce an intermediate heat map by: eliminating pixels in the heat map corresponding to the one or more reference roads in the road network map, eliminating background pixels corresponding to background of the heat map, and thinning the heat map; and determine start coordinate information and end coordinate information associated with the one or more target roads based on the intermediate heat map according to a linear transformation.

Abstract: Techniques for automatically detecting when a ride requester has requested a ride-share ride on behalf of a guest rider using some or all of the communications between the driver and ride requester are described herein. For example, a server can obtain chat logs between a ride requester and a driver and process the chat logs to identify whether the ride requester has requested a ride on behalf of a guest rider. In particular, the server can train an artificial intelligence model (e.g., a machine learning model) to predict potential guest rider behavior. Once trained, the server can obtain chat logs comprising chat messages sent between a driver and a ride requester, and apply a representation of the chat logs as an input to the trained artificial intelligence model to determine whether guest rider behavior is detected.

Abstract: The present disclosure relates to systems and methods for displaying one or more points of interest (POIs) of a location-based service (LBS). The method may include receiving a request for an LBS sent by a user via a user terminal and obtaining a start location based on the request for the LBS. The method may also include determining one or more recommended POIs based on the start location and POI data associated with the start location. Further, the method may also include transmitting signals to the user terminal to display at least one of the one or more recommended POIs.

Abstract: An automatic agent may be improved through feedback. A user input may be received through a user interface. A plurality of current utterance variables may be obtained by tokenizing the user input. The automatic agent may execute a machine learning policy to generate a reply to the user input based on the plurality of current utterance variables. A different reply may be obtained in response to an indication that the reply will lead to a breakdown, wherein the breakdown comprises an unhuman response from the automatic agent according to the machine learning policy. The machine learning policy may be adjusted based on the plurality of current utterance variables and the different reply.

Abstract: Systems and methods involving obtaining point cloud data from one or more sources corresponding to one or more zones within a real-world space, the point cloud data representing surface features of structures detected within the one or more zones; defining a plurality of first-level nodes based on the point cloud data, individual first-level nodes corresponding to obtained point cloud data corresponding to individual zones of the one or more zones; identifying connections between two or more first-level nodes, the connections between the two or more first-level nodes based on connections between the point cloud data for the zones corresponding to the two or more first-level nodes; and defining a plurality of second-level nodes, individual second-level nodes corresponding to aggregated subsets of first-level nodes for which connections are identified.

Abstract: Methods, systems, and apparatus for recommending alternative destinations in ride-sharing services are provided. A computing device implementing the method may start with receiving a trip request from a user device. The trip request may include an origin and a destination. Then the computing device classifies the trip request into one of a plurality of trip purpose categories based at least on the origin and the destination of the trip request, the rider's information, and a machine-learning classifier trained to predict the one trip purpose category of the trip request. In response to the one trip purpose category belonging to a preset group of trip purpose categories, the computing device determines one or more alternative destinations for the trip request, and sends to the user device, the one or more alternative destinations.