32results about How to "Improve conversation skills" patented technology

The invention discloses a session processing method and device, a terminal and a storage medium, and belongs to the technical field of the Internet. The method comprises the steps of: receiving session requests of first users, and determining user levels of the first users, wherein the user levels are used for indicating importance degrees of the first users, and the session requests are requestsfor carrying out sessions with a second user, which are initiated by the first users; according to the user levels of the first users, determining display information of session window tags of the first users, wherein the display information is used for indicating the user levels of the first users; and according to the display information of the session window tags of the first users, displayingthe session window tags of the first users on a display interface of the second user. By displaying the session window tags on the basis of distinguishment of the user levels, the second user can clearly identify the importance degrees of a plurality of first users, thereby avoiding a case that important users close session windows due to excessively long waiting time, and improving a session effect and session efficiency.

The invention discloses a tone recognition method and apparatus, an electronic device and a computer-readable storage medium. The method includes the steps of acquiring a sample text; acquiring a plurality of recording data for the sample text, the plurality of recording data being speeches obtained by reading the sample text by using a plurality of types of tone; extracting tone features in the plurality of recording data, and determining tone types corresponding to the tone features; and establishing a tone recognition model based on the sample text, the tone features in the plurality of recording data, and the corresponding tone types. The method can make each sample text be marked with the features of various types of tone and corresponding tone types so that in online application, themethod can perform speech recognition on user-entered speech data and also the tone recognition of speech data through the pre-established tone recognition model at the same time, thereby realizing the recognition of various types of tone in the speech data and improving the speech recognition effect.

The invention relates to a voice session method and device, a computer device and a storage medium. The method comprises the steps that a target session text of a to-be-generated voice is acquired; semantic recognition is performed on the target session text to obtain a semantic feature vector corresponding to a target word segment in the target session text, and an emotional rhythm label corresponding to the target word segment is determined based on the semantic feature vector corresponding to the target word segment; a semantic boundary in the target session text is determined, and a boundary text is inserted into the target session text based on the semantic boundary to obtain a first session text; the first session text and the emotional rhythm label corresponding to the target word segment are input into a voice synthesis model to synthesize a target session voice, wherein the voice synthesis model encodes the boundary text, and voice pause information in the target session voiceis obtained based on a vector obtained through encoding; and the target session voice is output. By the adoption of the method, the session effect can be improved.

Embodiments of the invention provide a session method based on companion equipment, equipment and a system, relates to the technical field of information, and mainly aims at solving defects in the prior art that companion equipment is poor in adaptability. The method comprises steps of receiving user feature information sent by companion equipment; obtaining session scene information corresponding to the user feature information, wherein different pieces of user feature information are corresponding to different pieces of session scene information; and determining the session scene information as session scene information of the companion equipment, and carrying out a session according to the session scene information when the companion equipment receives the session scene information. The invention is suitable for sessions based on companion equipment.

A children language ability evaluation method based on artificial intelligence comprises the following steps: 1) compiling a PYTHON crawler program to obtain required audio, question and answer corpora and encyclopedia knowledge, and merging special scene corpora accumulated in project practice in combination with optimism into an optimism corpus for calling; 2) carrying out model training on thecaptured voice material through a Deep Learning deep learning module, STT conversion can be carried out on the collected child voice, performing voice recognition through a preferred local area network offline STT, and besides, third-party voice recognition can also be adopted as a second option in an internet access mode. The invention discloses a language acquisition and analysis system. The system carries out noise reduction and voiceprint analysis on the collected audio data, and matches the audio data with the child to which the audio data belong. After audio-to-text processing is carriedout on audio data of children, analysis is carried out through an artificial intelligence system, and the leaping degree, the language logicality, the expression initiative, the gift degree and the like of speaking of the children are evaluated.

An embodiment of the invention provides a communication link establishment method, a communication link establishment device, equipment and a storage medium. The communication link establishment method comprises the following steps that the server forwards a session invitation initiated by a first client to a second client, and after a response of accepting the session invitation returned by the second client is received, a channel token is generated and transmitted to the first client and the second client respectively, and the first client and the second client can simultaneously initiate arequest of entering a channel and acquire audio and video data of an opposite terminal by communicating with sound network equipment. According to the communication link establishment method providedby the embodiment of the invention, the delay of the communication link can be reduced, and the audio and video live broadcast session effect is improved.

The invention relates to the technical field of hearing aids, in particular to a hearing aid system for caring old people, which integrates the functions of basic hearing aid, regular care reminding, real-time analysis and the like, and can improve the session communication ability between the old people and other people under the condition of not increasing the physical and mental burden of the old people. Negative emotions such as loneliness and self-abstraction caused by hearing impairment of the elderly are avoided, and great convenience is provided for the hearing-impaired elderly; on the other hand, voice care prompting can be carried out by recording the sound of relatives of the elderly, and the use experience of the elderly is further improved; more importantly, the hearing aid system also has the function of paying attention to and analyzing the voice and mood dynamics of the old in real time, and can perform corresponding analysis, so that family members of the old can be conveniently helped to accurately master the physical and psychological states of the old in real time, and the physical and psychological health of the old can be regularly evaluated according to an analysis report; and physical and psychological diseases of the old are further prevented.

The invention discloses a response processing method, intelligent equipment and a computer readable storage medium. The following steps are included: when the intelligent equipment receives current input information of a user, dialogue information including the current input information of the user being firstly analyzed; obtaining the current intention of the user and user basic information at least including problem solving conditions or user emotions; taking the current intention of the user and the basic information of the user as input of a dialogue prediction model to perform model prediction to obtain a model prediction result; and when the model prediction result is used for representing ending of the dialogue, feeding back a response language for guiding ending of the dialogue tothe user.

The embodiment of the invention provides a user simulator construction method and device, electronic equipment and a storage medium, and the method comprises the steps: obtaining a first dialogue state representation of a user simulator and a second dialogue state representation of a dialogue robot; obtaining a current round of dialogue state representation according to the first dialogue state representation and the second dialogue state representation; predicting a dialogue target of the current dialogue based on the current dialogue state representation; and generating a reply statement ofthe current round of dialogue based on the dialogue target. The user simulator constructed by the embodiment of the invention is favorable for improving the generalization of the user simulator so asto better simulate the dialogue situation of a real user, thereby improving the dialogue effect of the user simulator.

The invention provides a semantic fusion method and system based on a micro-service, and a medium, and relates to the technical field of man-machine interaction. The method comprises the following steps: S1, constructing a plurality of business skill semantic understanding micro-services according to business demands; S2, taking a user corpus text as input of a plurality of business skill semantic understanding micro-services to obtain a plurality of semantic understanding results; and S3, conducting semantic decision making on the multiple semantic understanding results through confidence levels, and acquiring an optimal semantic result at the same time. According to the method, a plurality of semantic understanding function modules can be fused by adopting a tree structure, the problems that a large memory is occupied when a plurality of business skill semantic understanding models are loaded, different versions of the same installation package are incompatible and service engineering is not facilitated are solved, dialogue performance is improved, and meanwhile, good user experience of man-machine dialogue is guaranteed.

The invention provides an intelligent robot for adjuvant therapy of autistic children. The intelligent robot comprises a vision recognition module and an intelligent question and answer module, wherein on-site finding, moving and rotary finding are adopted in the search path to achieve 360-degree full-view search; target color recognizing is carried out on the basis of YUV color space; for targettracking, a target is always achieved in the view center algorithm, a method of cloud voice recognition is used, an NAO robot obtains voice questions of a user, the voice questions are uploaded to a web cloud to recognize, and return recognition characters are subjected to corresponding question analyzing and answer searching processing; the emotion of the autistic children is recognized, and corresponding motion is executed; and voice of the autistic children is recognized, and corresponding motion is executed. The intelligent robot has the beneficial effects that a music bank with rich content and various types is constructed, the conversation ability of the autistic children is reinforced through intelligent chat, and therefore the therapy effect of the autistic children is improved remarkably.

The invention provides a dialogue processing method and device, equipment and a storage medium. Relates to the technical field of computer application, in particular to the artificial intelligence fields of intelligent question answering, intelligent dialogue, intelligent customer service and the like. According to the specific implementation scheme, a first answer is obtained; obtaining dialogue related data corresponding to the first answer; under the condition that the dialogue related data meets a first intervention condition, determining an intervention mode of the first answer; performing intervention processing on the first answer based on the intervention mode to obtain a second answer; and outputting the second answer. According to the technical scheme, personalized answers can be generated, and the dialogue ability of the dialogue system is improved.

The invention discloses an intelligent toy. The intelligent toy comprises a toy bear body. A speaker box, a battery pack, a circuit board and a remote control unit are arranged in a body of the toy bear body. The circuit board is electrically connected with the speaker box and the battery pack. The remote control unit is used for controlling the speaker box to give out different kinds of voice or conversations. The speaker box is located on the head of the toy bear body. A switch is arranged on an arm of the toy bear body. The circuit board is provided with a master control circuit module, a USB circuit module, a speaker circuit module, a switch control circuit, a test circuit module and an oscillating circuit module. The intelligent toy is diversified in function, multiple modes for child learning exist, the learning initiative of children can be improved, the varieties of child learning can be increased, the children can learn voice, listen to music and conversations and the like while playing the toy, and the language competence and conversation capacity of the child are improved.

The disclosed voter with touch screen comprises a voting unit. Wherein, the voting unit includes a CPU connected to all other units, a display unit, a communication unit, and a power. This invention has compact structure and fit to wide application ac simple operation.