487 results about "Human computer dialogue" patented technology

The invention concerns a human-machine dialog system (1) comprising:a support (3) having a plurality of identical docking stations (30), each docking station (30) being associated with a universal human-machine dialog device (4), each universal human-machine dialog device (4) comprising at least a display member (41) and a sensor member (40),a plurality of modular members (2), each modular member (2) being arranged to be positioned in a docking station (30) in a removable and interchangeable manner and comprising a human-machine dialog interface (20) arranged to cooperate with said display member (41) and / or said sensor member (40).

The invention discloses a voice interaction method. The voice interaction method comprises the steps that current voice information input by a user is sent to a voice recognition server and a character string obtained by the voice recognition server through conversion of the current voice information is received; whether the character string is in accord with a current conversation scene or not is judged according to key information words, extracted from context in the conversation process of a user and stored, of the current conversation scene; if the character string is in accord with the current conversation scene, the character string is matched according to preset matching rules and the key information words; the matched character string is sent to a semantic parsing server, and after an operation instruction returned by the semantic parsing server according to the character string is received, the operation instruction is executed and an answer is made to the user. The invention further discloses a corresponding system and an interaction terminal. By the adoption of the scheme, the sense of experience of man-machine conversation is improved, user-friendly service is provided for the user, and user-friendly man-machine interaction is realized.

The invention relates to a method for changing a plane video to a stereoscopic video, based on real time human-computer conversation, which belongs to the multimedia technical field of computers. The method comprises that: a computer divides a whole video sequence into sub-sequences that have related contents; a user designates any frame of each sub-sequence as a key frame; the computer carries out wave-filtration to all plane video frames to enhance image edge information and sharpen the edge of the foreground objects of the video frame; foreground object segmentation is carried out to the key frames and non-key frames and profile curves and depth maps are extracted until depth map sequences corresponding to all the frames of the original plane video sequence are generated; the smoothed depth map sequences are utilized for rendering and generating multi-visual angle views corresponding to each moment, and the views are synthesized into stereoscopic video frames; the stereoscopic video frames of all the moments are composed into the stereoscopic video sequences. As the method is based on the real time human-computer conversation, the precise depth map of each frame can be obtained, thus well realizing the changing from the plane video to the stereoscopic video and finally obtaining the stereoscopic video with better effect.

The invention belongs to the field of computer vision and the field of man-machine interaction and provides a personal character and interpersonal relation identification-based man-machine interaction system and a work method. The system comprises a detection module, an information feature extraction module, an information feature integration module, a character and preference determining module, a relation determining module and a human-computer dialogue module, wherein the detection module is used for detecting pictures or videos, and real zones of human face and human body images can be obtained; the information feature extraction module is used for extracting human face and human body image features and analyzing human face characters and human body behaviors; the information feature integration module is used for combining human face and human body information output via a former module, and feature vectors for each person are formed; the character and preference determining module is used for analyzing personal characters and preference according to the input feature vectors; the relation determining module is used for determining relations between people in the images or the videos, and the human-computer dialogue module is used for choosing a topic for communication with a user according to the personal characters and preference and the relations among the people. The system is used for analyzing the personal characters and preferences of the people and the relations among the people in the pictures and the videos, and a suitable topic is chosen for taking initiative to communicate with the user; a human-machine interaction mode is enabled to be intelligent, natural and human-friendly.

The invention discloses a natural language understanding method based on a pre-training model. The method comprises the following steps: establishing a pre-training model based on a bidirectional depth Transformer; carrying out word segmentation processing on the to-be-understood statement, and respectively adding special tags at the beginning and the end of the to-be-understood statement to obtain a text vector of the to-be-understood statement; taking the text vector of the to-be-understood statement as input, and calling the pre-training model to obtain a text semantic vector of the to-be-understood statement; performing intention recognition; and performing entity identification. According to the method, intentions can be accurately and comprehensively understood, entities can be recognized, and a solid foundation is provided for subsequent conversations. The quality and the user experience of the man-machine conversation system can be remarkably improved.

The invention provides a knowledge retrieving method of an artificial intelligence dialogue and a knowledge base automatic improving method. The knowledge retrieving method of the artificial intelligence dialogue comprises the steps of 1, obtaining a question statement of a user; 2, conducting semantic and intention analysis on the question statement to obtain a question identification; 3, according to the question identification, searching for a local answer in a local knowledge base; 4, obtaining the local answer, and fusing local question data and replying the local question data to the user; 5, when the local answer is not searched in the local knowledge base, according to the question identification, searching for a remote answer in a data source; 6, obtaining the remote answer, and fusing the remote answer and replying the remote answer to the user. According to the knowledge retrieving method of the artificial intelligence dialogue and the knowledge base automatic improving method, when the local knowledge base has no resource, the remote answer in the data source is searched for, and the occurrence that because entity coverage of a mapping knowledge domain is not sufficient, during a human-computer dialogue, a robot cannot give an answer is avoided.

The invention is a design based on an intelligent design platform application system, belonging to the field of the application of the calculation analysis module of a power grid decision. The system adopts the hardware environment of a regional power grid dispatching decision support system, support platform software and data platform software, wherein the hardware part can be divided into computer network, data acquisition, real-time monitoring and control, historical data storage, network safety equipment, dispatcher workstation, maintenance workstation and the like, and the software of the system adopts a WINDOWS system, a database system Oracle, an advanced support platform and application software. Real-time data collected by a computer network and a data acquisition module is sent to a real-time monitoring and controlling module by being communicated with a communication server and is variedly processed and transformed; relevant parameters are stored in a historical data storage module; finally, a machine operating the advanced application software of an operation power grid serves as a terminal workstation which also doubles as a client; and a man-machine conversation window provides a handling means for operation personnel and dispatchers.

The invention provides an Internet of Things-oriented smart home service system, including an affair reminding module, an equipment management module, an equipment control module, a contextual model management module and a man-machine conversation service module which are connected with a central processing module. The equipment management module and the equipment control module are connected with smart home equipment through a wireless network. The Internet of Things-oriented smart home service system provided by the invention can use a smart phone or a tablet computer as an operating platform, performs unified control of networking smart home equipment through WiFi technology, can be connected with a home common sense database, and can interact with a user in a man-machine asking and answering mode; has contextual models, and can intelligently adjust and control a plurality of connecting devices in one time; and provides various intelligent presetting and adjustment functions according to specific functions and empirical data of the controlled devices, thereby achieving intelligent services.

The invention comprises a real-time traffic status speech query and broadcast method and system. The method includes the following steps that: traffic status query speech information is received in a vehicle; speech recognition is performed on the traffic status query speech information, and the traffic status query speech information is converted into query text information; semantic analysis is performed on the query text information, the semantic instruction of the query text information is determined; the semantic instruction is sent to a server; the server returns corresponding traffic status response text information according to the semantic instruction; and broadcasting is performed in the vehicle according to the traffic status response text information. With the method and system of the invention adopted, the potential safety hazards of traffic status query in a driving process can be eliminated; a user neither needs to look at a screen nor needs to perform manual operation, the user only needs to perform wake-up operation to carry out human-computer dialogue so as to query real-time traffic statuses, and therefore, more convenient and efficient services can be brought to the user.