26 results about "Audio signal flow" patented technology

The invention relates to the technical field of atmosphere lamp control, in particular to a dynamic atmosphere lamp control method and system based on music rhythms. The method comprises the following steps: collecting a real-time audio signal stream, carrying out multi-scale wavelet decomposition and audio rhythm hierarchical structure perception, and constructing an audio rhythm feature perception map; performing spatial audio azimuth analysis based on the real-time audio signal stream, and performing adaptive dynamic atmosphere lamp parameter control based on the audio rhythm feature perception map to obtain an adaptive dynamic atmosphere lamp control strategy; according to the self-adaptive dynamic atmosphere lamp control strategy, initial atmosphere lamp control is carried out, a real-time image in the area is obtained, weak light intensity area coverage compensation optimization is carried out, and a dynamic atmosphere lamp space lighting effect optimization strategy is constructed; personalized natural physiological rhythm mining and physiological-music resonance frequency fine tuning are carried out, and user physiological-music resonance frequency fine tuning parameters are generated. According to the invention, high-response atmosphere lamp control which dynamically adapts to the light effect preference change of the user is realized.

The invention belongs to the technical field of communication network management, and discloses a heterogeneous communication network convergence system and convergence method. Comprising the steps of collecting multi-mode audio signal data and performing data cleaning to obtain a standard audio signal stream; constructing an efficient transmission architecture to perform data transmission and encryption on the standard audio signal stream; decoding the standard audio signal stream to obtain an original audio signal stream; performing priority ranking on the original audio signal flow to obtain an authority audio signal sequence; constructing a communication network convergence framework to perform channel resource allocation on the permission audio signal sequence to obtain a channel matching audio signal stream; monitoring frame parameters of the communication network convergence frame in real time and performing dynamic resource optimization to obtain an optimized communication network convergence frame, and sending the optimized communication network convergence frame to a preset scheduling terminal; and efficient fusion of heterogeneous communication networks is realized by constructing an optimized communication network fusion framework.

The present application discloses an emotion recognition method, apparatus, storage medium and device. The method includes: first, obtaining a target video to be recognized, and separating an audio signal stream and an image signal stream therefrom; then, respectively inputting the audio signal stream and the image signal stream into a pre-constructed emotion recognition model to predict a first preliminary emotion recognition result and a second preliminary emotion recognition result corresponding to a target user in the target video; wherein, the emotion recognition model is obtained by alternately training using a sample audio signal stream and a sample image signal stream, and then training by means of a parameter update method of gradient fusion; then, determining a final emotion recognition result of the target user according to the first preliminary emotion recognition result and the second preliminary emotion recognition result. Since the present application first constructs an emotion recognition model by using a training method related to modalities and a parameter update method of gradient fusion, the accuracy of the recognition result can be improved when using this model for emotion recognition.

An audio streaming device for streaming an audio signal to at least one hearing assistance device (32). The audio streaming device comprises at least one input transducer (21), an electronic sensor device (28) adapted to sense a gravity acting on the audio streaming device (20), and a transmitter (30) adapted to stream audio from the at least one input transducer (21) to the at least one hearing assistance device (32). The audio streaming device (20) further comprises a shock protection unit (31) adapted to compare an output signal from the electronic sensor device (28) with a first threshold value and to stop streaming of an audio stream from the at least one input transducer (21) by the transmitter (30) when the output signal from the electronic sensor device (28) exceeds the first threshold value.

A triggered data streaming system captures audio from an aircraft and generates an audio signal. Once the system obtains a trigger indicating a potential emergency on the aircraft, the system streams the audio signal to an external destination outside of the aircraft. The trigger may be an external trigger received from outside the aircraft or an internal trigger generated from sensors within the aircraft.

Methods and systems capture sounds on multiple sensors of a device worn by a user to produce streams of audio signals as preliminary audio output, transform the streams of audio signals and extract features from a resulting transform of the audio signals, and denoise the preliminary audio output by inputting the extracted features to a statistical model or neural network. As a result, a processed signal having a higher SNR than the preliminary audio output is produced. A humanly perceptible output in the form of transcribed text or a voiced version of the user' speech may be generated from the processed signal. The wearable device may be an earbud having microphones dedicated to capture sound conducted through air in an environment in which the user is situated and speech of the user conducted through bone of the user to one of the sensors.

Communications system used for trading communications that interfaces with a transcription system to enable a service of transcribing trader communications. During an audio call, a stream of audio signals transmitted by a communications device via a user-selected communications line are received together with metadata from the communications device. The audio signals and metadata can be provided during the audio call and via an API gateway to a transcription system in one or more data formats that enable the transcription system to transcribe the stream of audio signals into text and tag the text with the metadata.

The invention relates to the technical field of wireless audio transmission, and discloses an audio sharing loudspeaker device matched with a heart sound stethoscope for teaching. The device comprises a parameter acquisition unit, a calibration unit, a clock unit, a distribution unit, a decoding unit, a feedback unit and an evaluation unit, wherein the parameter acquisition unit acquires environmental data and generates audio adjustment parameters; the calibration unit calibrates the teaching audio signal according to the audio adjustment parameter; the clock unit extracts time stamp information from the calibrated audio signal flow and realizes clock synchronization; the distribution unit generates and pushes multiple paths of audio distribution streams based on time synchronization, and monitors transmission interference and processes data packet loss at the same time; the decoding unit decodes the audio signal stream and adjusts the volume; the feedback unit collects audio playing effect feedback data and optimizes transmission; the evaluation unit evaluates audio distribution quality and reasonably switches channels. According to the invention, the problems of asynchronous playing and unstable sound quality in multi-terminal audio teaching are solved, and the synchronism and teaching quality of heart sound auscultation teaching are improved.

The application provides a system and method for real-time uplink and downlink audio anomaly detection. The real-time uplink and downlink audio anomaly detection method comprises: at the uplink device end, diagnosing audio problems of each key node of an audio uplink transmission link in a predetermined order; at the downlink device end, diagnosing audio problems by combining data packet processing and decoding anomaly monitoring. The diagnosis process triggers three independent detection paths, including device running state detection, audio signal stream detection and network transmission quality detection. Through diagnosis, the root cause of the audio problem can be determined, which can be used to provide guidance information for repairing the audio problem to the user.

A method for detection of a keyword in a continuous stream of audio signal, by using a dilated convolutional neural network, implemented by one or more computers embedded on a device, the dilated convolutional network comprising a plurality of dilation layers, including an input layer and an output layer, each layer of the plurality of dilation layers comprising gated activation units, and skip-connections to the output layer, the dilated convolutional network being configured to generate an output detection signal when a predetermined keyword is present in the continuous stream of audio signal, the generation of the output detection signal being based on a sequence of successive measurements provided to the input layer, each successive measurement of the sequence being measured on a corresponding frame from a sequence of successive frames extracted from the continuous stream of audio signal, at a plurality of successive time steps.

A noise cancellation technique is presented with improved frequency resolution. The method includes: acquiring a digitized noise signal from an environment in which the audio signal stream is present; receiving a data sample from the digitized noise signal; appending one or more additional samples to the data sample to form a series of samples, where magnitude for each of the one or more additional samples is substantially zero; computing a frequency domain representation of the series of samples in the frequency domain; shifting the frequency domain representation of the series of samples in time using the digital processor circuit, thereby producing a shifted frequency domain representation of the series of samples; converting the shifted frequency domain representation of the series of samples to time domain to form a portion of an anti-noise signal; and outputting the anti-noise signal into the audio signal stream to abate the noise through destructive interference.

The utility model discloses a two-way talkback processing system, and belongs to the field of voice signal processing. The system combines an advanced audio processing technology, and particularly uses an audio echo processing chip and an audio power amplifier module. Through an accurate echo cancellation algorithm (including frequency domain and time domain filtering), dynamic noise suppression and real-time gain adjustment, the system can significantly improve audio signal quality, avoid echo and self-excitation squeal phenomena, and ensure voice definition. The two-way audio signal stream design and the modular structure of the system enable the system to adapt to complex noise environments and various audio application scenarios, including talkback between drivers and passengers and processing of emergency alarm signals. In addition, through the design of audio bus transmission, audio power amplifier module amplification, transformer isolation and the like, stable transmission and high-quality output of audio signals are ensured. On the whole, the stability, definition and adaptability of the audio communication system are improved, and the echo problem in a traditional system is solved.

The invention relates to the technical field of audio control, in particular to a self-adaptive audio control method and system for a vehicle-mounted scene. The method comprises the following steps: acquiring an in-vehicle audio signal stream based on a vehicle-mounted microphone array, performing dynamic range compression and frequency response balance optimization, and constructing an in-vehicle reference audio stream; carrying out multi-position acoustic characteristic analysis according to the in-vehicle reference audio stream, carrying out acoustic transmission path distribution evolution, and constructing a vehicle cabin three-dimensional sound field model; extracting multi-dimensional driving behavior data, performing dynamic driving preference modeling, and constructing a driving behavior portrait; carrying out audio sensitivity calculation according to the driving behavior portrait, carrying out dynamic adaptive adjustment based on a vehicle cabin three-dimensional sound field model, and constructing a dynamic audio adjustment parameter; and performing safety optimization constraint and adaptive tone quality gain on the dynamic audio adjustment parameters, and outputting a final vehicle-mounted audio. The audio parameters are adjusted in real time based on the vehicle scene, and the driving safety and auditory experience are improved.

The invention relates to the technical field of human voice synthesis, and discloses an interactive human voice synthesis system based on a Frust webpage end, and the system comprises the steps: obtaining a user input code, generating an audio signal in real time according to the user input code, and playing the audio signal; user input is statistically analyzed according to historical data; executing the historical data prediction input strategy, and predicting codes possibly input by a user; executing an audio difference analysis and mistaken touch detection strategy, analyzing the difference between the audio signals, and starting a mistaken touch detection process; real-time feedback and personalized automatic repair suggestions are provided; executing the audio signal flow imaging strategy to generate a visual audio signal flow graph; codes and graphs are combined, and synthesis and modification of audio signals are completed through graphical interface operation; respectively generating corresponding audio signal flow diagrams and comparing the audio signal flow diagrams; after creation is completed, a final audio file is generated; according to the scheme, the audio creation efficiency is improved through the graphical interface, real-time audio synthesis and intelligent repair.

Systems, methods, and computer storage media are disclosed for transcript aggregation. In an example implementation, an audio mix processing map is generated representing an audio signal flow defined by a multi-track sequence that mixes audio sources arranged in audio tracks into a final mix. Using the audio mix processing map, an aggregated audio track is generated that represents which of the audio sources contribute to the final mix during which ranges of time. The aggregated audio track is transformed into an aggregated transcript track by associating transcript segments from transcripts of the audio sources with corresponding ranges of the ranges of time.

This invention relates to the field of audio control technology, and more particularly to an adaptive audio control method and system for in-vehicle scenarios. The method includes the following steps: acquiring in-vehicle audio signal streams based on an in-vehicle microphone array, performing dynamic range compression and frequency response balance optimization to construct an in-vehicle reference audio stream; analyzing the acoustic characteristics at multiple locations based on the in-vehicle reference audio stream, and performing acoustic transmission path distribution evolution to construct a three-dimensional sound field model of the vehicle cabin; extracting multi-dimensional driving behavior data, performing dynamic driving preference modeling, and constructing a driving behavior profile; calculating audio sensitivity based on the driving behavior profile, and dynamically adaptively adjusting it based on the three-dimensional sound field model of the vehicle cabin to construct dynamic audio adjustment parameters; applying safety optimization constraints and adaptive sound quality gain to the dynamic audio adjustment parameters, and outputting the final in-vehicle audio. This invention adjusts audio parameters in real time based on the vehicle scenario, improving driving safety and auditory experience.

Communications system used for trading communications that interfaces with a transcription system to enable a service of transcribing trader communications. During an audio call, a stream of audio signals transmitted by a communications device via a user-selected communications line are received together with metadata from the communications device. The audio signals and metadata can be provided during the audio call and via an API gateway to a transcription system in one or more data formats that enable the transcription system to transcribe the stream of audio signals into text and tag the text with the metadata.

An example immersive audio signal processing system and a computer-implemented method for generating a target arena environment audio stream are provided. The example immersive audio signal processing system includes a plurality of multi-lobe digital sound wave capture devices positioned within the arena environment. The plurality of multi-lobe digital sound wave capture devices is configured to direct first beamformed lobes to a playing region of the arena environment, second beamformed lobes to a spectator region of the arena environment, and third beamformed lobes to a noise source region of the arena environment. A digital signal processor is configured to isolate noise audio components originating from at least the spectator region or the noise source region from the audio signal stream and generate a target arena environment audio stream.

Systems and methods for real-time uplink and downlink audio abnormality detection are provided. In some embodiments, the methods and systems for real time uplink and downlink audio abnormality detection includes, at the uplink device, diagnosing audio issues at each major stage of an audio uplink pipeline in an ordered fashion and at the downlink device, diagnosing the audio issues by combining abnormality monitoring from packet handling and decoding. Diagnosing audio issues triggers three independent check paths, which include a device operation check, an audio signal flow check and a network transmission check. The diagnosis results in a root cause of the audio issues being discovered. This root cause may be used to supply a user with information to fix the audio issue.

An example immersive audio signal processing system and a computer-implemented method for generating a target arena environment audio stream are provided. The example immersive audio signal processing system includes a plurality of multi-lobe digital sound wave capture devices positioned within the arena environment. The plurality of multi-lobe digital sound wave capture devices is configured to direct first beamformed lobes to a playing region of the arena environment, second beamformed lobes to a spectator region of the arena environment, and third beamformed lobes to a noise source region of the arena environment. A digital signal processor is configured to isolate noise audio components originating from at least the spectator region or the noise source region from the audio signal stream and generate a target arena environment audio stream.

This invention provides a multi-source automatic gain control method, system, electronic device, and medium, belonging to the field of vehicle technology. The method includes: acquiring raw audio data from multiple audio sources; converting the raw audio data into an audio signal stream; analyzing the signal energy of the audio signal stream to obtain a current signal energy value; obtaining a gain adjustment coefficient based on the current signal energy value and a preset target level value; smoothing the gain adjustment coefficient to obtain a target gain control value; adjusting the gain of a digital signal processor according to the target gain control value; and outputting an audio signal with a target loudness. This invention eliminates the problem of sudden volume changes caused by inconsistent output levels of different audio sources by dynamically adjusting the gain in real time based on the signal energy value (RMS value) of the audio signal stream, thus maintaining a stable and consistent output loudness. It utilizes RMS value calculation, which is computationally efficient and has low resource overhead, as the core of its dynamic gain control, which operates in real time and does not significantly increase the system load.

Communications system used for trading communications that interfaces with a transcription system to enable a service of transcribing trader communications. During an audio call, a stream of audio signals transmitted by a communications device via a user-selected communications line are received together with metadata from the communications device. The audio signals and metadata can be provided during the audio call and via an API gateway to a transcription system in one or more data formats that enable the transcription system to transcribe the stream of audio signals into text and tag the text with the metadata.

A triggered data streaming system captures audio from an aircraft and generates an audio signal. Once the system obtains a trigger indicating a potential emergency on the aircraft, the system streams the audio signal to an external destination outside of the aircraft. The trigger may be an external trigger received from outside the aircraft or an internal trigger generated from sensors within the aircraft.

This invention discloses a deep learning-based method for detecting the arrival of Morse signals in an audio signal stream, serving as a pre-processing technique for automatic Morse signal decoding. The method uses 1 second of audio sampling data as a frame, performing a signal decision for each frame. The purpose of signal decision is to determine whether a given frame is a Morse signal. This includes a data preprocessing module that performs noise reduction and feature extraction on a single frame; and a deep learning decision module that inputs the preprocessed features into a deep neural network and determines whether the frame is a Morse signal based on the network's output. Based on the decision results for each frame, the arrival and end times of the Morse signal are determined. This invention can efficiently and accurately detect the arrival of Morse signals, a prerequisite for automatic Morse signal decoding. Furthermore, this invention has good adaptability to low signal-to-noise ratio and fading conditions, and has broad practical value.

Audio apparatus and methods of operating the same produce high-fidelity audio signals by leveraging discrete electronic devices each including a receiver, a microphone or microphone array, and a casing housing the receiver and the microphone or microphone array. The electronic devices are temporarily fixed to a stand in static positions as spaced apart from each. Sound from the ambient environment outside the stand is captured using the microphones or microphone arrays of the electronic devices and converted to respective streams of audio signals. The streams of audio signals are digitally processed by a processing unit, on board the stand, the electronic devices or an external device, of the apparatus based on information including distance between and orientations of the electronic devices in their static positions. The processed audio signals are transmitted digitally or as audio from one or more speakers.

The invention relates to a speech translation method and device, equipment and a medium in the technical field of the Internet, and the method comprises the steps: responding to a speech translation event, continuously receiving an original audio signal flow corresponding to the event, and determining a plurality of voice-containing frame sequences corresponding to the original audio signal flow; for each human voice-containing frame sequence, performing non-human voice noise reduction processing on the human voice-containing frame sequence to obtain a pure human voice frame sequence; reasoning a to-be-translated text sequence corresponding to the pure human voice frame sequence by adopting a voice recognition model; and translating the to-be-translated text sequence into a target text sequence according to a target language corresponding to the speech translation event. The method can efficiently and accurately translate the voice in a streaming manner, and greatly improves the translation robustness in the acoustic environment of noise interference and mixed sound sources.