50 results about "Voice engine" patented technology

A computer-implemented method for operating an artificial intelligence voice agent system includes receiving voice input through communication channels; analyzing converted text through natural language processing (NLP) pipelines implementing intent recognition and sentiment analysis detecting emotional cues using a multimodal large language model (LLM); generating response content using machine learning models trained on domain-specific corpora; converting generated responses to synthetic speech through text-to-speech (TTS) engines; integrating with a customer relationship management (CRM) platforms or an enterprise resource planning (ERP) database; and implementing continuous learning by updating language understanding models using conversation logs, voice recognition parameters based on user feedback, and response generation patterns. One implementation is a computer-implemented system and method that operates a suite of intelligent interactive devices and platforms including an artificial intelligence voice agent, enhanced communication platforms, an intimacy companion system, and augmented / virtual reality eyeglasses. Further, one implementation includes AR / VR eyeglasses that project visual content onto interchangeable lenses or directly onto the user's retina via laser-based retinal projection, provide prescription adjustments, incorporate ear-mounted sensors for monitoring physiological parameters like heart rate, oxygen saturation, and blood pressure, and utilize wireless data transmission, onboard environmental sensing, and remote calibration, all designed to offer dynamically adaptive, secure, and context-aware interactions across communication, personal assistance, health monitoring, and immersive augmented or virtual reality environments.

A system and method of real-time feedback confirmation to solicit a virtual assistant response from an evolving semantic state of at least a portion of an utterance. A user accesses a virtual assistant on an electronic device having the system and / or method configured to capture a command, a question, and / or a fulfillment request from audio such as, the speech emitted from the speaking user. The speech may be intercepted by a speech engine configured to transcribe the speech into text that is matched with the fragment pattern's regular expression to generate a fragment and / or the speech may be processed with a machine learning model to identify fragments. The fragments are identified by a domain handler configured to update a data structure of the current semantic state of the utterance in real-time on an interface of an electronic device.

Methods, systems, and computer storage media for providing speech synthesis using a code-mixed speech engine in a speech synthesis system. A code-mixed speech engine supports generating natural and intelligible speech in a target speaker voice—for code-mixed-text of two or more languages—based on a code-mixed speech model that supports both code-mixing and cross-locale voice transfer scenarios. In operation, code-mixed training data associated with a plurality of different languages is accessed. A code-mixed speech model—associated with a training engine and an inference engine that support generating code-mixed synthesized speech—is generated. The code-mixed speech model is deployed. A request being received for synthesized speech of a speech synthesis service. An instance of code-mixed synthesized speech is generated. The instance of code-mixed synthesized speech is generated using the code-mixed speech model. The instance of code-mixed synthesized speech is communicated for output on an interface associated with the speech synthesis service.

The invention is suitable for the technical field of intelligent door locks, and provides a voice interaction processing method, device and system, an intelligent door lock and a cloud server, and the method comprises the steps: receiving a voice instruction sent by a user, and carrying out the local recognition, so as to obtain a local recognition result and the confidence of the local recognition result; when the local recognition result is matched with the local instruction set and the confidence is high, executing a local operation corresponding to the local recognition result; otherwise, establishing a secure communication session with the cloud server based on voiceprint biological characteristics extracted from the voice instruction; sending the voice instruction to a cloud server through the secure communication session, and receiving a cloud operation instruction and / or a dynamic response text returned by the cloud server; and executing corresponding operation according to the cloud operation instruction, and / or converting the dynamic response text into voice for broadcasting through a text-to-voice engine. The intelligent door lock voice interaction system solves the problem that the real-time performance, the reliability, the safety and the interaction intellectualization are difficult to balance in the existing intelligent door lock voice interaction technology.

Embodiments include a computing system, computer-implemented method and non-transitory computer readable medium for predicting events and voice-AI warnings. According to embodiments, data corresponding to a predicted event is received, and users that are predicted to be affected by the predicted event are identified. A voice-AI engine is initiated to perform a voice-AI call to the identified users, where the voice-AI call provides a warning to each of the users.

The invention provides a voice call real-time transcription system and method, and relates to the technical field of computers, and the system comprises a network element module which is used for obtaining corresponding audio data when a call request of a user side is detected; the voice streaming engine is used for carrying out hierarchical compression on the audio data based on a preset perceptual weighted vector quantization algorithm to obtain audio compressed data, and carrying out format conversion processing on the audio compressed data to obtain temporary audio data; the voice engine is used for performing feature extraction on the temporary audio data to obtain multi-modal feature data, and processing the multi-modal feature data based on a preset voice recognition model to obtain text information; and the analysis and optimization module is used for obtaining corresponding real-time transliteration text data according to the text information and a preset vocabulary library based on a preset large model. According to the method, the voice information is comprehensively represented by using the multi-modal feature data, so that the voice recognition model can more accurately convert the voice into the text.

A system and method for automating sales and support interactions using computerized agents that are customized based on attributes of a target customer / buyer is disclosed. An artificial intelligence salesperson (AIS) utility / module collects / retrieves activities and attributes information about a target customer to generate a custom AIS that can be used to communicate on a sales call or a technical support call with the customer. The AIS is designed to substantially emulate a real person to better communicate with the target customer. The AIS can change on the fly for each caller, learn from sighs audible expressions of the target customer, and react to emotional changes of target customers. To create the AIS, the AI module modifies an existing speech pattern, and / or creates its own speech patterns in real time, before passing the new and / or modified scripts to the voice engine, which then “speaks” on a call to the target customer.

The invention provides a vehicle-mounted voice engine system and an interaction method thereof, a vehicle and electronic equipment, and the system comprises a first information interaction module group which is used for converting first voice information inputted by a voice receiving module into voice-text information, and transmitting the voice-text information to a target application program; the second information interaction module group is used for receiving feedback information output by the target application program according to the voice-text information and converting the feedback information into first text-voice information; and the voice output module is used for performing voice broadcast according to the first text-voice information. The technical problem solved by the invention is that a vehicle-mounted voice engine system (SDS) is deeply connected with a pre-installed voice assistant App, so that in the process of calling various pre-installed voice assistant App interfaces, the vehicle-mounted voice engine system needs to determine the return value of the interface to confirm the information broadcasted to a user, but the broadcast information is realized by the pre-installed voice assistant App, and the user experience is improved. Therefore, the vehicle-mounted voice engine system repeatedly develops the broadcasting function of the part, and resources are wasted.

A computer system and method are configured to permit a single agent to simultaneously interact with multiple customers. The computer system includes a call center server configured to communicate with (a) a plurality of customer devices, and (b) an agent device. The call center server is further configured to permit the agent device to communicate simultaneously with the plurality of customer devices. A text-to-speech (TTS) engine may be used converts text entered into the agent device into speech that is transmitted by the call center server to at least some of the plurality of customer devices.

The invention discloses a real-time translation interaction method, system and device based on a double-sided screen device, and the method comprises the steps: translating source language content displayed on a main screen into a target language in real time through a translation server, and then displaying the target language on the main screen, an auxiliary screen or a projection screen; monitoring a control instruction based on the main screen and the auxiliary screen in real time, and synchronously displaying the control content displayed on one screen for the control object in a corresponding area of the same control object on the other screen; according to a trigger request of the voice translation touch button, a voice signal is obtained and translated in real time to obtain target language content, and the target language content is converted into voice through a TTS voice engine and played; a signal source direction is determined by using a time delay estimation algorithm, and environmental noise compensation is carried out by using an adaptive noise compensation algorithm. According to the technical scheme, flexible and convenient multi-mode translation interaction display is realized, the translation communication efficiency is remarkably improved, and the translation interaction experience in a complex scene is improved.

The application discloses a kind of virtual standardization patient image generation and dialogue method and system, comprising: using large language model to extract the multidimensional information of patient from original medical data and construct structured case data, infer the visual features of patient, and integrate visual features as portrait description prompt word;Using text encoder, portrait description prompt word is converted into high-dimensional semantic vector, guide latent diffusion model to generate patient portrait image in line with medical data;Convert user voice signal into natural language text as role prompt word;According to the structured case data of patient and preset emotional factor, system prompt word is constructed;Using large language model and based on role and system prompt word, the reply text of user voice signal is generated;Start voice engine to convert reply text into reply voice, and output reply voice through loudspeaker.The application can generate virtual patient entity with visual, audible and intelligent interaction capability according to static medical text.

The invention relates to an industrial intelligent ERP (Enterprise Resource Planning) production management system and a key technology, belongs to the technical field of industrial informatization software, artificial intelligence voice interaction and equipment predictive maintenance, and aims at solving the problems that a traditional industrial ERP system is insufficient in flexibility, workshop voice interaction is difficult, and equipment maintenance is passive and low in efficiency. The core innovation of the invention lies in that: a module hot plug architecture is adopted, so that a user can switch a full-function ERP system / a lightweight chemical single system / a purchase-sale-stock combined system within 5 seconds; according to an anti-noise voice engine, the instruction recognition rate gt in the 90dB workshop environment is determined; 92%, response delay lt; 200 ms; a closed loop is predictively maintained, an equipment health model is called in real time in work order execution, and shutdown loss is reduced by 40%. The problems that a traditional ERP system cannot be flexibly split, the workshop voice interaction rate is low, and equipment maintenance is passively responded are solved.

A text-to-speech engine creates audio output that includes synthesized speech and one or more media content item snippets. The input text is obtained and partitioned into text sets. A track having lyrics that match a part of one of the text sets is identified. The location of the track's audio that contains the lyric is extracted based on forced alignment data. The extracted audio is combined with synthesized speech corresponding to the remainder of the input text to form audio output.

In one aspect, an example method includes (i) obtaining, by a computing system, structured data; (ii) generating, by the computing system using a natural language generator, a textual description of the structured data; (iii) transforming, by the computing system using a text-to-speech engine, the textual description of the structured data into synthesized speech; and (iv) generating, by the computing system using the synthesized speech, a synthetic video comprising the synthesized speech.

A peer-to-peer conversation captioning (“PPCC”) system includes a first electronic device for a first user, a second electronic device for a second user, and a PPCC server. The PPCC system further includes a virtual relay entity, which includes a first virtual router for the first language, a second virtual router for the second language, and a global switch. The virtual routers include a speech-to-text engine, a translation engine, a text-to-speech engine, and a multiplexer. The speech-to-text engine converts a user's spoken language to texts and sends the texts to the global switch. Then, the translation engine receives and translates the texts where the text-to-speech engine converts the translated texts to voice. The multiplexer combines the translated texts and the voice to be sent to another user.

The invention provides a text-to-voice method, electronic equipment, a readable storage medium and a computer program product. The text-to-voice method disclosed by the invention comprises the following steps: adding text-to-voice tasks into a task queue, and sorting the text-to-voice tasks in the task queue according to a timestamp sequence; in response to a task acquisition request of a consumer thread, sending the first text-to-voice task in the task queue to the consumer thread; in response to a task processing request of the consumer thread, sending a text-to-voice task obtained by the consumer thread to a text-to-voice engine; acquiring voice data returned by the text-to-voice engine; wherein the text-to-voice engine is used for converting text data into voice data.

A system and method for automatically generating a report based on cockpit voice recognition is provided. The system includes a data acquisition module, a speech engine, and a report generator. The data acquisition module is configured to acquire cockpit voice data and vehicle parameter data. The speech engine is configured to convert the cockpit voice data into formatted text data. The speech engine includes: a pre-processing module configured to convert the cockpit voice data into formatted voice data; an ASR module configured to convert the formatted voice data into text data using automatic speech recognition; and a post-processing module configured to reorganize the text data into formatted text data using text processing. The report generator is configured to generate a vehicle standard operating procedure (SOP) compliance report based on the formatted text data and a report template. The vehicle SOP compliance report indicates how the investigated behavior derived from the cockpit voice data complies with the standard operating procedures defined by the report template.

A human-to-machine transmission system and a machine-to-human transmission system includes a non-linear encoder that converts data representing an utterance into a frequency spectrum representing an aural range. An encoder in the human-to-machine transmission system that converts the output of the non-linear decoder into a bitstream and a decoder converts the bitstream into the frequency spectrum. An automatic speech recognition engine of the human-to-machine transmission system processes an output of the decoder to deliver an output. The machine-to-human transmission system includes a language model that replies to inquiries and a neural network that converts the text-to-speech received from a text-to-speech engine. It also includes a decoder that converts an input into the frequency spectrum, a vocoder that converts the frequency spectrum into audio frames, and a loudspeaker that converts the audio frame into audible sound.

In one embodiment, a method includes receiving, from a client system associated with a first user, a first audio input. The method includes generating, based on a plurality of automatic speech recognition (ASR) engines, a plurality of transcriptions corresponding to the first audio input. Each ASR engine is associated with a respective domain of a plurality of domains. The method includes determining, for each transcription, a combination of one or more intents and one or more slots associated with the transcription. The method includes selecting, by a meta speech engine, one or more combinations of intents and slots associated with the first user input from the plurality of combinations. The method includes generating a response to the first audio input based on the selected combination and sending, to the client system, instructions for presenting the response to the first audio input.

A real-time feedback confirmation system and method for requesting a virtual assistant response from the evolving semantic state of at least a portion of an utterance. A user accesses a virtual assistant on an electronic device having a system and / or method configured to capture commands, questions, and / or requests from audio, such as speech, uttered by a speaking user. The speech may be captured by a speech engine configured to transcribe the speech into text that is matched with regular expressions of fragment patterns to generate fragments, and / or the speech may be processed by a machine learning model to identify fragments. The fragments are identified by a domain handler configured to update a data structure of the current semantic state of the utterance in real time on the interface of the electronic device.

The invention discloses a multi-mode fusion interaction system of a self-service terminal, and the system comprises a voice interaction module which is used for recognizing a voice instruction based on an AI voice engine and converting the voice instruction into a first operation command; the gesture recognition module is used for capturing a hand action by using a 3D camera and converting the hand action into a second operation instruction; the touch screen interaction module is used for processing touch input on terminal touch screen equipment as a third operation instruction; the fusion processing unit is used for fusing, analyzing and sorting the first operation command, the second operation command and the third operation command to determine a final operation command; the output module is used for executing corresponding operation according to the final operation instruction; according to the method, multiple interaction modes can be integrated, the interaction flexibility is greatly improved, the user can freely switch among different interaction modes according to own habits, the current environment and the physical condition, and the convenience and ability of special groups to use various devices are improved.

A computer system and method are configured to permit a single agent to simultaneously interact with multiple customers. The computer system includes a call center server configured to communicate with a plurality of customer devices, and an agent device. A text-to-speech (TTS) engine converts text entered into the agent device into speech that is transmitted by the call center server to at least some of the plurality of customer devices. Convert SMS emails and chats messages from agent to customer devices. Pause conversation if detected that agent needs more time. Fill pauses and interruptions of the agent with a voice filler.

A language-independent dictionary-trained grapheme-to-phoneme converter and text-to-speech engine for improved speech recognition is disclosed. Techniques are described for recognizing a spoken wakeup word (WW) or command using a speech recognition system that does not need to be trained with any speech data that matches the WW / command for a human machine interface. Techniques train a word segmentation device and include decomposing a word into a plurality of combinations by splitting the word from a database at a plurality of different points for each of the plurality of combinations of unique sub-words. The word includes one or more writing units and one or more corresponding acoustic units. Techniques map acoustic units constituting a word to writing units constituting the word to generate an acoustic unit-to-writing unit mapping. The technique includes assigning a subset of the acoustic units to each of the unique sub-words based on the acoustic unit-to-write unit map to generate an acoustic unit-to-sub-word assignment of the word. Techniques accumulate acoustic units to sub-word assignments for a plurality of words from a database to create a sub-word likelihood dictionary.

The application discloses a power supply and energy efficiency service interpretation system based on AI artificial intelligence, a background management module of which selects corresponding voice broadcast services and energy use suggestion services according to user demand information; when the background management module selects the voice broadcast services, a voice broadcast service module encodes power supply and energy efficiency service voice messages to be executed in a queue to obtain corresponding audio stream data, and a text-to-speech engine performs voice synthesis output; when the background management module selects the energy use suggestion services, an energy use suggestion module queries corresponding cases in a power supply and energy efficiency service case database through an energy use recommendation label, adds a name value corresponding to case information to a content header sent to a browser according to an http-equiv attribute of the case information, and then jumps to a specified address. The application realizes the intelligentization of customer service and improves customer service experience.

A system is provided.SOLUTION: A system comprising: means for receiving a voice input of a user; means for converting the voice input into digital data; means for transmitting the digital data to a server; means including a speech recognition engine for converting the digital data into text data; means including a text-to-speech engine for analyzing the text data, understanding a user's intention, and acquiring information from an external service based on the user's intention; means for generating a response message in a natural language based on the acquired information; means for converting the response message into voice data; and means for transmitting the voice data to a terminal.SELECTED DRAWING: Figure 1

The application belongs to the technical field of voice instructions, and specifically discloses a general voice instruction distribution method, an electronic device and a storage medium. The method of the application realizes the API by accessing a voice engine, converts returned semantic callback instructions into unified entity classes, defines two parameter variables, inputs the unified parameters into a semantic executor interface, instantiates the semantic executor interface through dynamic proxy, and the semantic executor interface obtains the callback interface modified by annotation, the semantic callback instruction method and the unified parameters of the semantic callback instruction method through reflection. The parameter variables in the unified parameters are matched with the callback interface modified by annotation and the parameters of the annotation, and it is judged whether to execute. If it is judged to be callback, the callback semantic instruction is executed. Otherwise, it is prompted that the command is not supported. When the voice engine is replaced, the new engine API can be flexibly accessed, the voice instructions of different voice engines are distributed and processed, and the research and development cost is reduced.

Disclosed are systems and methods for enabling collaborative AI-human voice interactions in an outbound call environment. An AI voice engine synthesizes real-time speech responses, optionally using a voice model representing a specific human agent. A sentiment analysis engine monitors user speech to classify emotional state, and an emotion modulation engine dynamically adjusts tone, pitch, and prosody of AI output based on inferred sentiment or operator commands. An operator console allows live human oversight, including editing AI-generated text, selecting emotional tone presets, or overriding responses entirely. The system supports real-time disclosure of AI identity and records call metadata for compliance. Training modules log operator interventions, sentiment patterns, and interaction outcomes to improve future AI behavior, enabling a scalable voice platform that balances automation with human empathy.

A system for improving human performance includes a software application with an onboarding engine, an initialization engine, an optimization engine, and a voice engine to populate a user profile, and use the profile to select and optimize a training protocol. The application is executed on a computer with processor, memory, user interface, and storage. The application exchanges information with the user and accesses a file containing the user profile. A method for improving human performance includes onboarding a new user by dynamically generating questions to create a profile; selecting an initial training protocol for the user based on the profile and verified training pathways; modifying the initial protocol based on user feedback to create an optimized training protocol, and guiding the user's training session with voice instructions that are generated by a voice engine.

The application provides a voice call real-time transcription system and method, and relates to the technical field of computers.The system comprises a network element module for acquiring corresponding audio data when detecting a call request of a user terminal; a voice stream sending engine for performing hierarchical compression on the audio data based on a preset perceptual weighting vector quantization algorithm to obtain audio compression data, and performing format conversion processing on the audio compression data to obtain temporary audio data; a voice engine for performing feature extraction on the temporary audio data to obtain multimodal feature data, and processing the multimodal feature data based on a preset voice recognition model to obtain text information; and an analysis optimization module for obtaining corresponding real-time transcription text data based on a preset large model and according to the text information and a preset vocabulary library.The application comprehensively represents voice information by utilizing multimodal feature data, so that the voice recognition model can more accurately convert voice into text.

The application discloses a kind of intelligent voice engine capability sharing model establishment and evaluation method and system include, obtain the audio data generated by dispatcher in the dispatching process when power system operation, and the data is preprocessed;The data after pre-processing is divided into first data set and second data set, first data set is used to establish voice engine sharing model, and second data set is used to verify voice engine sharing model;According to the verification result, the performance of voice engine sharing model is graded evaluation.The audio data of dispatching command platform is converted into structured index information, through data intelligent analysis, help accident analysis and disposal, combined with the actual scene of dispatching production intelligent learning and optimization are carried out, improve the accuracy of intelligent voice analysis, realize the scientific evaluation of dispatching language specification, improve the quality of dispatching command.