166 results about "Latency (audio)" patented technology

The invention relates to the technical field of audio playing optimization, and discloses a low-delay multi-room audio playing optimization method and system, a storage medium and equipment, and the method comprises the steps: building a global time synchronization system based on a precision clock protocol in a playing equipment cluster, and enabling all playing equipment to operate based on a unified time reference; monitoring network state parameters in real time, dynamically calculating and adjusting the audio buffer depth of each playing device based on the monitored network quality, and obtaining audio data in advance according to the playing progress and the network state; generating a global playing timestamp according to the global time synchronization system, and controlling all playing devices to perform synchronous starting and progress control of audio playing based on the global playing timestamp; a transmission strategy of audio data is dynamically optimized based on real-time network state parameters, packet loss recovery and bandwidth adaptation are achieved, sub-millisecond multi-room audio synchronous playing is achieved, and meanwhile performance parameters can be automatically optimized according to different network environments and hardware configurations.

The invention relates to the technical field of simultaneous interpretation, and discloses a simultaneous interpretation data processing method and system based on a POE microphone array. The method comprises the following steps: synchronously acquiring multi-language original audio streams and meeting place environment noise spectrum features through a distributed microphone array powered by the Ethernet; after time domain framing is carried out on the audio stream, adaptive filtering is carried out by using a dynamic noise reduction weight coefficient to obtain a primary pure voice segment; dividing the multi-language speech endpoint detection model into independent speech units with language labels through a pre-trained multi-language speech endpoint detection model, and matching a corresponding acoustic model to generate a phoneme-level time alignment sequence; comparing and outputting a term replacement instruction stream in real time in combination with a simultaneous transfer term library, and generating an intermediate semantic representation vector after fusion; and the low-delay encoder converts the voice parameter sequence into a target language voice parameter sequence, and drives the waveform synthesizer to generate final simultaneous transmission audio. The method optimizes the whole process processing, gives consideration to the simultaneous transmission accuracy and real-time performance, and is suitable for a multilingual meeting place scene.

The embodiment of the invention discloses a display device and a screen projection display method, and the method comprises the steps: after receiving a screen projection connection instruction sent by a terminal device, sending a starting notice to a player through a screen projection service, and enabling the player to build a data pipeline based on the starting notice. And controlling the player to receive the code stream data sent by the terminal equipment after the data pipeline is established, and processing the code stream data by using the data pipeline when the player receives the code stream data to obtain a video picture and an audio. The display displays a video picture, and the audio output device plays audio. According to the embodiment of the invention, data transmission is directly carried out through the terminal equipment and the player, so that communication delay caused by inter-process communication is reduced, the received code stream data is directly processed after the player establishes the data pipeline, cache is reduced, and the delay time of screen projection display is shortened.

The invention provides a model compression and data enhancement fused lightweight deep forged voice detection method. The method comprises the steps of obtaining and processing a public voice data set and a large-scale self-supervision pre-training voice model; performing structured pruning and knowledge distillation to obtain a lightweight voice model; performing audio preprocessing and diversified data enhancement on the true and false voice samples to obtain an enhanced true and false voice data set; performing faking task joint fine tuning on the lightweight voice model to obtain a lightweight deep faking voice detection model; and locally deploying the model to obtain a localized counterfeit voice detection system, and carrying out real-time authenticity judgment. According to the method, calculation complexity and reasoning time delay are remarkably reduced, local deployment is carried out on a resource-limited end side platform, detection generalization and robustness are improved, and low-delay, low-power-consumption and high-robustness detection performance is achieved.

A system for latency-aware orchestration and performance optimization in AI-driven telephone communication, consisting of: a speech capture unit configured to capture an analog audio signal from a telephone interface and convert the analog audio signal into a digital audio signal stream; a feature extraction unit that is operationally coupled with the speech acquisition unit and is configured to generate a feature representation of the digital audio signal stream through spectral decomposition, noise reduction, and temporal segmentation; an AI inference processor communicatively connected to the feature extraction unit, configured to run one or more AI models for automatic speech recognition, natural language understanding, and emotion recognition on the feature representation to generate intermediate results for inference; a latency orchestration controller coupled to the AI ​​inference processor, wherein the latency orchestration controller is configured to monitor latency across multiple processing stages, predict cumulative delay propagation using a hybrid latency estimation model, and orchestrate the execution scheduling of the AI ​​inference processor based on the predicted latency deviation; a performance optimization unit coupled with the latency orchestration controller and configured to dynamically adjust computational accuracy, inference batch size, and feature processing resolution based on latency thresholds and quality constraints set by the latency orchestration controller; and a transmission synchronization array configured to time-align the processed output generated by the AI ​​inference processor and transmit it to a remote communication node, with the transmission synchronization array maintaining deterministic time coordination between successive packets and the orchestrated inference results.

The invention relates to the technical field of elderly care, solves the problems that in the prior art, due to the fact that noisy and intermittent elderly voice is lack of robust sentence bound recognition and interruption repair, misinformation and missing information of calling-for-help triggering are prone to occurring, and time delay is high, and provides an audio processing method and device for elderly calling-for-help, equipment and a medium. The method comprises the following steps: preprocessing a voice signal of the elderly to obtain a preprocessed voice signal; performing statement boundary recognition and labeling on the preprocessed voice signal to obtain a voice fragment set; carrying out interruption segment and silent interval detection on the voice segment set to obtain an initial voice segment sequence, and carrying out intelligent recombination and time sequence repair on the initial voice segment sequence to obtain a target voice segment sequence; and carrying out emergency call recognition on the target voice segment sequence to obtain an emergency call judgment result. According to the invention, the accuracy of distress call identification of the elderly is improved.

The disclosure describes systems and methods for improving latency for wireless speakers. The system can establish a buffer pipe between a wireless chip of a media player and a wireless chip of a wireless speaker using an inter-IC sound protocol. The system can receive samples of audio data through the media player. The system can communicate the audio data to the wireless speaker using the buffer pipe to bypass the transport layer stack of the media player and the wireless speaker. The system can cause a transmitter of the wireless speaker to provide the audio data to a digital audio converter for output to a speaker of the wireless speaker.

The invention discloses a WebM protocol low-delay video and audio translation and subtitle optimization method and system, and belongs to the technical field of data processing, and the method comprises the steps: analyzing a WebM audio and video stream; establishing a video track-audio track association list and constructing a dynamic causal map; generating an initial translation text based on the audio frame sequence, the video frame sequence and the dynamic causal atlas, extracting lip movement and action semantic data of video frames through a reverse generation model to generate a completion translation text, and embedding SimpleBlock elements; defining a subtitle core gene, a non-core gene and a position gene according to the code rate data, the complete translation text and a user visual attention thermodynamic diagram, dynamically cutting the non-core gene based on code rate fluctuation, and adjusting the position gene in combination with the thermodynamic diagram to generate adaptive subtitle data; by synchronously playing and collecting feedback data, the edge weight and the subtitle position gene of the dynamic causal atlas are optimized. According to the invention, the robustness of audio and video translation and the self-adaptability of subtitle display are realized.

The invention discloses an ultra-low delay live broadcast method, system and device based on edge calculation, and a medium, and relates to the technical field of real-time audio and video interaction. The method comprises the steps that a target edge computing node receives an original media stream; monitoring an uplink network link and a downlink network link of the anchor end and the target edge computing node in real time, and determining a network link quality parameter; acquiring a historical network link quality parameter, and setting a link degradation standard and a link excellent standard according to the historical network link quality parameter; determining a target compensation strategy according to the network link quality parameter, a first preset condition and a second preset condition; and the target edge computing node processes the original media stream by adopting the target compensation strategy to generate a compensated media stream, and forwards the compensated media stream to a receiving end. By adopting the scheme, the overall performance of ultra-low delay live broadcast can be improved.

Low latency, over-the-top (OTT), and / or adaptive bitrate (ABR) content streaming is provided. Content delivery is enhanced by determining if a fragment of a content segment at a content delivery network (CDN) edge node meets a threshold for preferential encapsulation and transport. If met, preferential encapsulation and transport to the client device is provided; otherwise, it defaults to non-preferential encapsulation. The size of the fragment is quantified at a parser of the CDN edge node or an ABR segment encryption system. The ABR system may be connected between a content source and a CDN origin and may include an encryptor that sends CMAF video and audio segment's fragment byte offsets metadata. Also, the CDN edge node may include the ABR system and an encryptor that sends an encrypted CMAF segment's fragment size to a threshold calculator of an HTTP server. Related apparatuses, devices, techniques, and articles are also described.

A telecommunications switch-type platform positioned between a private-branch exchange and a data network dynamically hosts artificial-intelligence agents that interpret, transcribe, translate and transliterate live call traffic while the session is in progress. Machine-readable instructions stored in local memory instantiate, migrate and terminate the agents on demand across dedicated processing resources such as field-programmable gate arrays, application-specific integrated circuits or system-on-chip devices, thereby maintaining end-to-end latency below 250 milliseconds. An integrated call-audio recorder captures bidirectional media streams for secure archiving without interrupting service. The platform exposes a network interface that passes both signalling and media traffic, enabling seamless deployment inside call-centre or enterprise environments and compatibility with softswitch architectures conforming to class-4 Computational Interpretation of Communication using Artificial Intelligence standards. The same functional stack is deliverable as a computer-implemented method and as a non-transitory machine-readable medium storing the instructions executed by the processing resources.

Disclosed are a method, system, and apparatus of a real-time multilingual transcription system and method. In one embodiment, a method includes continuously capturing an audio data and segment it into short segments; implementing a pre-trained enterprise-grade voice activity detection (“VAD”) system on each of the short segmental and filtering out non-speech segments to reduce computational waste, focusing resources on relevant audio data and minimizing latency. If speech is detected, a particular short segment is added to a processing queue. If speech is not detected, declining to add the particular segment to the processing queue, thereby reducing unnecessary processing.

Various implementations include processing, at a client device, an instance of audio data capturing a user voice query using an automatic speech recognition model to generate a sequence of instances of tokenizable query text. In many implementations, one or more instances of the sequence can be transmitted to a remote computing system prior to generating the entire sequence. In a variety of implementations, each instance in the sequence can be processed using a generative model which includes a streaming multi-head attention portion. Responsive output can be transmitted from the remote computing system to the client device, where the client device renders the responsive output to the user. In many implementations, the time between the user speaking the user query and the client device rendering the responsive output is reduced, thus decreasing latency in the system.

Described embodiments provide systems and methods for latency optimization for cloud applications. An agent of a client device comprising an audio decoder and a video decoder can monitor video and audio data paths of an application communicating audio / video (A / V) data from one or more servers to the client device. The agent can measure, using the audio decoder and the video decoder, an A / V latency and a lip-sync status of the video and audio data paths of the application. The agent can determine, based on at least one or more measurements of the A / V latency and the lip-sync status, to enable a low latency mode for at least one of the video decoder or the audio decoder. The agent can configure, responsive to the determination, the low latency mode on one of the video decoder or the audio decoder.

This invention discloses a low-latency audio and video interactive transmission method supporting multi-terminal synchronization, belonging to the field of audio and video interactive transmission technology. It includes: enabling data communication between the client and server to achieve audio and video interactive data transmission according to the WebSocket communication protocol; adaptively selecting the optimal streaming media transmission channel based on the ICE algorithm; transmitting audio and video interactive data from the client to the server according to the optimal streaming media transmission channel; monitoring the transmission status of the audio and video interactive data in real time; and dynamically adjusting and optimizing the audio and video interactive data transmission based on the monitoring results. This invention solves the problems of existing methods that cannot effectively achieve low-latency transmission and cannot achieve high-quality real-time audio and video interaction across multiple terminals. This invention can ensure synchronous low-latency transmission of audio and video interactive data, providing users with a smooth and natural interactive experience, and enabling high-quality real-time audio and video interaction across multiple terminals.

The invention discloses a GIS equipment defect intelligent detection method based on computer vision, and aims to solve the problems that transient defect acquisition in a GIS cabinet is delayed and asynchronous alignment of audio and video is difficult. According to the method, time reference and annular pre-buffering are unified, collection of a streaming voiceprint trigger linkage event camera and a high-speed polarization camera, monotonic alignment Transform of arrival time difference prior constraint and residual time offset calibration, secondary collection closed loop of confidence gating and multi-modal fusion judgment are carried out, and multi-modal fusion judgment is carried out. The technical effects of low-delay triggering, order-preserving alignment, accurate time positioning, evidence chain integrity rate improvement and missing detection and false detection reduction are realized.

A class-4 telecommunications switch hosts artificial-intelligence agents that convert live speech to text, translate the text between languages, and synthesize natural speech in real time. The switch proxies calls among public trunks, PBX / media gateways, and cloud ACD / CRM services, embedding diacritic-rich transcripts and user-specific language-model personalization. Deployable at the customer edge, in the PSTN core, or as SaaS, the system supports one-to-one, one-to-many, many-to-one, and many-to-many call patterns. FPGA, ASIC, or SoC accelerators minimize latency and bandwidth, cutting capital cost while improving global voice interoperability and cybersecurity.

The invention relates to the technical field of communication, discloses a mining intelligent voice and video interactive communication method and system, and aims to solve the problems of poor audio and video signal quality, low bandwidth utilization rate, lack of semantic understanding and interaction lag in a complex mine environment. The method comprises the following steps: acquiring an original signal through an intrinsic safety type audio and video terminal; voice enhancement is carried out by adopting sound source orientation estimation and a deep complex network; reconstructing a high dynamic range video based on the physical imaging model; realizing audio and video semantic alignment and key event extraction by using a lightweight spatial-temporal feature fusion network; the bandwidth is dynamically allocated by combining the event confidence coefficient and the channel state, and high / common priority data is transmitted in a graded manner; and ground and underground real-time communication is realized through a bidirectional interaction channel. According to the invention, the signal availability, the bandwidth efficiency and the emergency response capability are remarkably improved, and high-stability, low-delay and high-fidelity communication is guaranteed.

A system for synchronizing and spatializing media, comprising client device and server device. A method for operation includes: tracking the real-time spatial coordinates of both devices via wireless sensors; synchronizing media playback using a timestamp-based protocol that compensates for network latency and clock offset; and applying directional audio filters, such as Head-Related Transfer Functions (HRTF), on the client device to render audio appearing to originate from server device's physical location. The method is further characterized by compensating for acoustic propagation and signal processing delays to ensure accurate spatiotemporal alignment. The system may be enhanced by a multimodal Al configured to perform operations such as real-time voice-preserving translation and the generation of synchronized, spatialized haptic feedback based on semantic event markers.

Audio latency measurement is provided. A device is configured to play, via a speaker, an audio wave in storage accessible by the device. The device is configured to detect the played audio wave by a microphone, the device to identify a time of detection of the audio wave. The device is configured to play, via the speaker, the played audio wave received by the microphone. The device is configured to detect the second played audio wave as being received by the microphone, the device to identify a second detection of the audio wave. The device is configured to determine a latency of communications of the device based on a difference between the first detection of the played audio wave and the second detection of the second played audio wave. The device is configured to communicate the latency to one of a second device or a user interface.

Apparatuses, systems, and methods for low-latency audio-to-face animation with emotion detection are disclosed herein. The system may receive a first audio stream associated with a first device and a second audio stream associated with a second device, and provide, concurrently, a first segment of the first audio stream and a second segment of the second audio stream as inputs to an emotion detection artificial intelligence (AI) model to obtain first emotion data and second emotion data. The system may then provide, concurrently, a third segment of the first audio stream with the first emotion data and a fourth segment of the second audio stream with the second emotion data as inputs to a face animation AI model to obtain first face pose data and second face pose data, and provide the first face pose data to the first device and the second face pose data to the second device.

This invention discloses a low-power edge-side simultaneous interpretation system and method based on audio-text synchronization and visual feature fusion, relating to the fields of multimodal human-computer interaction and machine translation technology. The system achieves high-precision audio-text timing alignment through an audio-text synchronization matching module, generates a lightweight visual feature stream using a visual feature processing module, and performs spatiotemporal fusion by a multimodal fusion inference module. Combined with edge-side heterogeneous computing power scheduling and dynamic power consumption control, it significantly reduces the power consumption of edge devices while ensuring low-latency translation of ≤20ms per frame. This invention fills the technological gap in edge-side low-power multimodal simultaneous interpretation and can be widely applied in edge scenarios such as mobile office, international communication, and smart wearables.

A system and method of generating a perceptual reflex encryption key (PRE-Key) may include the following steps: 1) delivering a spatially modulated auditory stimulus from a mobile device to a human subject via a secure audio output interface; 2) capturing, using a MEMS sensor subsystem, an involuntary physical response of the human subject to the auditory stimulus; 3) determining a response latency Δt between stimulus delivery and the captured response; 4) extracting a perceptual feature vector based on neocortical response approximations; and 5) computing the PRE-Key by hashing a combination of the auditory stimulus parameters, the perceptual feature vector, the physical response, and the response latency.

An ear-associated assistive system is disclosed that integrates acoustic sensing and inertial sensing to generate motion-compensated spatial parameters for hearing assistance and, in certain embodiments, to control stimulation for implantable auditory and / or vestibular interfaces. The system maintains temporal alignment between inertial samples and audio samples by maintaining a deterministic mapping between inertial sample times and audio sample indices, including across power-state transitions, thereby enabling reliable sensor fusion and consistent outputs. In certain implementations, an ear-frame coordinate system is established based on fixed mechanical placement of an inertial sensing subsystem relative to one or more microphones, and calibration parameters are stored to align sensor axes, microphone geometry and latency. A processor computes a head-motion state from inertial data and transforms an ear-frame direction estimate derived from acoustic data into a stabilized direction parameter expressed in a stabilized coordinate frame, optionally outputting a quality metric indicative of validity. The stabilized direction parameter and / or quality metric may be used for beamforming, binaural rendering and routing. The system may further detect motion events and apply safety gating rules to mitigate motion artifacts and constrain acoustic output and / or stimulation, subject to safety constraints and, in certain embodiments, clinician-defined bounds. Interoperability with an external directional accessory is also described, wherein inertial-acoustic fusion is used to stabilize directional operation and maintain consistent routing.

Synchronized output of audio on a group of devices can comprise sending audio data from an audio distribution master device to one or more slave devices in the group. Scores can be assigned to respective audio playback devices, the scores being indicative of a performance level of the respective audio playback devices acting as a master device. The device with the highest score is designated as a candidate master device and one or more remaining devices are designated as a candidate slave(s). A throughput test is conducted with the highest scoring device acting as the candidate master device. The results of the throughput test are used to determine a master device for a group of devices. Latency of the throughput test can be reduced by using a prescribed time period for completion of the throughput test, and / or by selecting a first group configuration to pass the throughput test.

This invention provides an amplifier with improved bias tracking performance based on a DSP chip. It includes a file acquisition module that directly acquires the audio file to be played via an interface. The audio file is pre-input into the DSP for standard waveform analysis, which speeds up processing and reduces latency during actual playback. During actual playback, the actual played energy is compared with the pre-calculated energy to determine if it matches the pre-designed output. If it matches, the pre-designed output is used directly; otherwise, it is further input into a model for processing, ensuring operational stability and safety.

A system includes a web application server, a plurality of hub servers, and a first audio streaming device. The first audio streaming device is configured to ping each hub server of the plurality of hub servers to determine a respective latency between the first audio streaming device and each hub server to provide a first plurality of latencies. The first audio streaming device is configured to transmit the first plurality of latencies to the web application server. The web application server is configured to store the first plurality of latencies in a latency table comprising a respective second plurality of latencies for a second audio streaming device. The web application server is configured to determine a compatibility for audio streaming between the first audio streaming device and the second audio streaming device based on the latency table.

This disclosure presents embodiments of a model acceleration method, a video generation method, an apparatus, a device, a medium, and a product. One specific implementation of the method includes: determining a sample set and an original video generation model, wherein the samples in the sample set include reference images and audio; performing conditional distillation on the original video generation model based on the sample set to obtain a first video generation model, wherein the number of forward propagation steps of the first video generation model is less than the number of forward propagation steps of the original video generation model; and performing distribution matching distillation on the first video generation model based on the sample set to obtain a high-speed video generation model, wherein the number of inference steps of the high-speed video generation model is less than the number of inference steps of the original video generation model. This implementation relates to audio-driven video generation, reducing the computational complexity and inference latency of the model, and facilitating deployment and use in scenarios with limited computing resources.

This application discloses a simultaneous interpretation method, apparatus, device, and storage medium, relating to the field of audio processing technology. The aforementioned simultaneous interpretation method is applied to a simultaneous interpretation device, which includes a locally deployed AI service module. The method includes: acquiring first audio data in a first language; converting the first audio data into second audio data in a second language offline, based on the locally deployed AI service module; and transmitting the second audio data via wired transmission to a first communication device in a call state, wherein the second audio data is provided to a second communication device engaged in a call with the first communication device. This method enables low-latency transmission during simultaneous interpretation.

Audio communication methods, devices, and systems, are provided with error correction overwrite for audio artifact reduction. One illustrative low-latency audio streaming method includes: receiving packets of digital audio data; applying an error correction code decoder to obtain a data stream that includes error-corrected data samples; providing a correction-limited data stream by replacing any of the error-corrected data samples that are outliers; and converting the correction-limited data stream into an audio signal.