528 results about "Audio mixing" patented technology

The invention belongs to the technical field of speech translation, and particularly relates to a video speech automatic translation method based on a pre-training model, and the method comprises the following steps: 1, preprocessing video and audio data; step 2, voice recognition and language detection; step 3, machine translation and text post-processing; step 4, speech synthesis and audio mixing; step 5, synchronizing video processing and subtitles; step 6, quality control and multi-dimensional evaluation; 7, carrying out model iteration and data closed loop; and step 8, system deployment and engineering implementation. Through deep fusion of efficient transfer learning of the pre-training model and the multi-modal technology, a high-precision, low-cost and easy-to-expand video speech translation solution is constructed, the time and labor cost of globalized content production is greatly reduced, the cross-language communication efficiency is improved, immersive multi-language experience is provided, and the method is suitable for popularization and application. And a data-driven continuous optimization mechanism is established, so that the system performance is improved along with the increase of the use scale.

The invention provides a method and a system for hierarchical audio classification. For getting high accuracy prediction with high resolution and low predictor complexity, the disclosed method uses a hierarchical classification approach with stateful prediction per frame aided by parallel AI transient detector for resetting the states of all stages at class transitions. To improve accuracy perfectly tagged database by innovative techniques of labeling are utilized. Further data augmentation is also done using signal processing techniques like audio mixing, blending of different type of data. The disclosed method applies short term audio normalization on database for normalized training and prediction of AI based Long Short-Term Memory (LSTM) networks. The disclosed method then uses a novel hierarchical classification approach with stateful LSTM prediction per frame aided by a parallel transient detector for resetting the states of all stages of hierarchical LSTM classifiers at class transitions.

The embodiment of the invention provides an audio processing method and device, electronic equipment and a storage medium. The method comprises the following steps: determining a target pulse response of a first audio signal, wherein the target pulse response is a stereo pulse response used for simulating sound reflection and attenuation effects of the first audio signal in an acoustic space; performing convolution processing on the first audio signal and the target pulse response to obtain a second audio signal; and performing audio mixing on the first audio signal and the second audio signal to output a third audio signal. According to the scheme provided by the invention, the stereo pulse response data capable of simulating the reflection and attenuation effects of the sound in the space can be synchronously determined when reverberation is carried out on the audio signal, and reverberation is carried out on the audio signal by using the stereo pulse response data, so that the sound is heard to be generated in different space environments; the reverberation effect of the audio signal is more vivid, the sense of space and the sense of layering of the sound are increased, and the sound is fuller and stereoscopic.

The invention relates to the technical field of audio processing and loudspeaker box control technologies, in particular to a loudspeaker box control method, system and device based on a built-in sound source and a storage medium, and the method comprises the steps: obtaining a selection instruction of a user, and calling a corresponding sound source file based on the selection instruction; receiving the called sound source file by using a built-in preset audio processing algorithm, and adjusting audio parameters of the sound source file; performing deep optimization combination on the processed audio parameters corresponding to the plurality of audio source files subjected to audio parameter adjustment by using an audio mixing algorithm so as to obtain a playing audio signal; and transmitting the playing audio signal to an integrated amplification circuit for power amplification, and enhancing the signal strength to drive a loudspeaker to work. The application has the effect of improving the audio quality and the user experience.

The invention discloses an audio data processing method and device, a storage medium and electronic equipment. The method comprises the steps that to-be-processed audio and at least one track point are acquired, initial audio data are generated based on the to-be-processed audio and the track point, loudspeaker gain corresponding to a target virtual loudspeaker is determined based on the initial audio data, and position information of the track point is used for determining a virtual source position in a virtual playing scene. Virtual loudspeakers are deployed in the virtual playing scene, the target virtual loudspeaker is at least one of the virtual loudspeakers, the target virtual loudspeaker represents the virtual loudspeaker determined based on the virtual source position, and the loudspeaker gain is used for simulating playing of the to-be-processed audio at the virtual source position; and performing sound mixing processing on the audio to be processed and the media information indicated by the source media information identifier based on the loudspeaker gain to generate target audio data. The technical problem that the playing effect of the audio data is poor is solved.

This invention provides a method and system for hierarchical audio classification, aimed at achieving high accuracy in prediction with enhanced decision time resolution. The disclosed method uses a perfectly tagged database which utilizes innovative techniques of labelling. Further data augmentation is also done using signal processing techniques like audio mixing and blending of different types of data. The disclosed method applies short term audio normalization on database for normalized training and prediction using AI based Long Short-Term Memory (LSTM) networks. The method and system employ the LSTM networks in a hierarchical structure to classify audio into desired 3 or more audio classes which include at least a background noise audio class. Decision time accuracy is improved by running the LSTM predictors over time overlapped slices and by using a separate transition detection neural network.

The invention relates to the technical field of audio processing, and discloses a multi-mode audio processing method and system based on an SOC. The method comprises the following steps: inputting multiple paths of audio input signals into an ARM processor of an SOC chip for processing, and generating a preprocessed audio data stream; inputting the preprocessed audio data stream into a basic audio processing stream executed by a digital signal processor (DSP) and a multi-modal feature extraction stream executed by an artificial intelligence (AI) accelerator in parallel; performing adaptive noise reduction processing and multi-channel audio signal sound mixing processing in the basic audio processing stream to generate a basic processing feature stream; performing multi-dimensional feature extraction and scene type identification in the multi-modal feature extraction stream to generate a high-level feature stream; according to the method, the processing capability of the SOC chip on different types of audio signals is improved, the continuity and smoothness of the reconstructed audio signals are ensured, and signal distortion is effectively avoided.

Methods for generating object-based audio programs that are renderable in a personalizable manner and include a speaker channel bed that is renderable without selection of other program content (e.g., to provide a default full-range audio experience). Other embodiments include steps of delivering, decoding, and / or rendering such programs. Rendering of the content of the bed or selected mix of other content of the program can provide an immersive experience. The program can include a plurality of object channels (e.g., object channels indicative of user-selectable and user-configurable objects), a speaker channel bed, and other speaker channels. Another aspect is an audio processing unit (e.g., an encoder or decoder) configured to perform any of the embodiments of the method or that includes a buffer memory storing at least one frame (or other segment) of an object-based audio program (or bitstream thereof) generated according to any of the embodiments of the method.

The invention discloses a real-time accompaniment tuning method and device, a storage medium and a program product. The method comprises the following steps: in response to a playing instruction for a target song, obtaining musical instrument audio data of each musical instrument corresponding to the target song; the musical instrument audio data is obtained by performing sound source separation on a music file of the target song according to musical instrument types; obtaining an adjustment strategy of a user for a target musical instrument in the target song in real time; performing adjustment processing on the musical instrument audio data corresponding to the target musical instrument according to the adjustment strategy to obtain adjusted musical instrument audio data; and based on the adjusted musical instrument audio data and the musical instrument audio data corresponding to the other musical instruments, obtaining customized accompaniment audio corresponding to the music file through sound mixing, and playing the customized accompaniment audio. Customized accompaniment audio meeting user requirements can be generated in real time.

The present invention provides a call system, a host device, a computer program product, a mobile terminal, and a cloud server. The call system (10) comprises a host device (11), a cloud server (12), and a plurality of clients (13). The host device (11) is configured to receive a first voice signal sent by a plurality of terminal devices (14), receive a second voice signal forwarded by the cloud server (12) and sent by the plurality of clients (13), perform audio mixing processing on the received first voice signal and the second voice signal, respectively generate a mixed audio signal corresponding to one of the terminal devices (14), send the mixed audio signal to the terminal device (14), generate a mixed audio signal corresponding to one of the clients (13), and forward the mixed audio signal to the client (13) by means of the cloud server (12). The call system may allow a user who is not within a coverage range of the wireless internal call system to access the call system for communication.

An audio infusion system and method are disclosed. A source audio track is separated into a plurality of audio tracks (e.g., instrumental, vocal, or mixes thereof) and the audio tracks are individually processed to generate a plurality of binaural beat tracks. At least one spatialized track is also generated by filtering the source audio track to provide a filtered track, generating one or more spatialization trajectories based on certain audio feature(s) of the source audio track (e.g., tempo) and a target end-state effect, and spatializing the filtered track using the spatialization trajectories. Other tracks may also be generated, such as one or more infrasonic tracks, ultrasonic tracks, enhanced bass tracks, and / or subharmonic tracks. The tracks may be played simultaneously or mixed for delivery to an end user device.

The invention discloses a bad video classification method and system based on multi-modal alignment and category balance, and the method comprises the steps: inputting a data set into a bad video classification model, training the bad video classification model, and enabling the bad video classification model to be called a model for short; when the total loss function value of the model does not decrease any more, a trained bad video classification model is obtained; in the training process, the model firstly extracts visual mixed features and audio mixed features from a video; the total loss function of the model comprises the weighted summation result of a first sub-loss function, a second sub-loss function and a third sub-loss function, and the first sub-loss function refers to the difference between the Gaussian distribution of the visual mixed features and the Gaussian distribution of the audio mixed features; the second sub-loss function refers to the difference between the visual mixing feature and the audio mixing feature; and the third sub-loss function is used for increasing the minimum distance between a model decision boundary and a sample point and reducing the influence of class imbalance on the training process.

The invention discloses a display device and an audio control method, and relates to the field of display devices, the audio control method is applied to the display device, the display device comprises a display, an audio output interface and a controller, the controller is in communication connection with the display and the audio output interface, and the audio output interface is configured to be in communication connection with a plurality of target audio devices. The target audio equipment is used for outputting audio data; the method further comprises the following steps: under multi-window display, firstly stopping sound mixing processing on each piece of audio data; and according to the determined target relationship, establishing an audio output channel for audio data transmission between each target window and the target audio equipment, so as to realize transmission of the audio data of different windows to different audio equipment, thereby improving the audio-visual experience of a user under multi-window display.

The invention relates to the technical field of audio processing, in particular to an airborne high-reliability intelligent voice call squelch reduction method, which comprises the following steps: performing ADC conversion on an acquired analog audio signal to obtain a digital audio signal; carrying out framing and FFT conversion on the digital audio signal, carrying out phase optimization, calculating a dynamic spectrum entropy weight, combining an auditory masking compensation factor, carrying out spectrum synthesis, and carrying out IFFT conversion and overlap elimination to obtain an audio signal after sound mixing processing; noise reduction processing is carried out on the audio signal after sound mixing processing through multi-mode feature extraction, and an audio signal after noise reduction processing is obtained; performing squelch judgment of Bayesian decision optimization on the audio signal subjected to noise reduction processing through energy feature extraction to obtain an audio signal subjected to squelch processing; performing automatic gain processing on the audio signal after the squelch processing to obtain an audio signal after automatic gain; and performing DAC conversion on the audio signal after automatic gain, and outputting a noise-reduced analog audio signal.

The invention discloses a sound mixing and noise reduction method and system, and the method comprises the steps: firstly, carrying out the recognition of the emotional state of a person in a vehicle through an emotion recognition method; and then, whether the person in the vehicle has a bad emotion which can influence the behavior state or not can be judged according to the emotion recognition result. And if the person in the vehicle does not have the impatient emotion, the current noise in the vehicle does not need to be processed. On the contrary, if the person in the vehicle has a restless emotion, the sound signal in the vehicle is immediately collected through the sound pick-up. And finally, a sound wave identification method can be adopted to extract a noise feature vector of a noise sound segment which affects the emotion of people in the vehicle from the collected sound signals, and a corresponding sound mixing sound wave band is generated based on the extracted noise feature vector and is fused with the noise sound segment which affects the people in the vehicle, so that the sound mixing sound wave band becomes a sound segment which can be accepted by people. Therefore, the influence of noise on the behavior state of people is reduced.

The invention relates to the technical field of audio signal separation, and provides an audio signal separation method and device, electronic equipment and a storage medium, and the method comprises the steps: obtaining frequency domain signals corresponding to audio signals of different channels of audio equipment; decomposing the frequency domain signal to obtain a coherent sound signal and an environment sound signal of each channel; under the condition that the energy corresponding to the environmental sound signals of the different channels is the same, determining a target environmental sound signal of each channel according to the phase difference between the environmental sound signals of the different channels; and separating the target environment sound signal of the corresponding channel from the audio signal of each channel, and obtaining the target coherent sound signal of the channel. According to the method, the device, the electronic equipment and the storage medium provided by the invention, secondary processing is carried out on the separated target coherent sound signal and the target environment sound signal by a sound mixer or a multi-channel algorithm, an auditory atmosphere with a surrounding sense and a presence sense is created, and the auditory experience of a user is improved.

An audio processing method includes establishing a wireless connection between a wireless headset and an electronic device, and establishing a wireless connection between the wireless headset and a charging case. The charging case collects a sound signal, generates first audio data, and sends the first audio data to the wireless headset. The wireless headset performs audio effect processing on the first audio data to obtain second audio data. The electronic device sends third audio data to the wireless headset. The wireless headset performs audio mixing processing on the second audio data and the third audio data to obtain fourth audio data and plays the fourth audio data.

Aspects of the present disclosure provide techniques, including devices and systems implementing the techniques, to output an adjusted audio mix to a wearable audio device in real-time using at least one pre-trained machined learned (ML) models. The pre-trained ML models enable adjustment of audio output with ultra-low latency. One example technique for adjusting audio output generally includes detecting, by one or more feedforward microphones, one or more spatialized sounds proximate a user, separating, using at least one pre-trained ML model, at least a portion of the spatialized sounds, wherein the separation is based on content of the sound, processing at least a portion of the spatialized sounds, and outputting, by the audio output device, an adaptive mix of the processed spatialized sounds. Such techniques may enable a user to filter out or reduce unwanted noises based on their sound content.

The invention provides a stereo synthesis method and device, a model construction method and device, equipment and a storage medium. The audio processing model comprises a sound event positioning and classification network and a sound source separation network. The stereo synthesis method comprises the following steps: acquiring a to-be-processed audio stream; inputting the to-be-processed audio stream into an audio processing model, and performing sound event classification and sound source orientation estimation on the audio stream by the sound event positioning and classification network to obtain the category and orientation of at least one sound event, the sound source separation network performs separation processing on the mixed audio according to the category and the orientation of the at least one sound event to obtain an audio signal of each sound event, and outputs the orientation and the audio signal of each sound event in the at least one sound event; and performing sound mixing processing on the audio signals corresponding to the at least one sound event according to the orientation of each sound event in the at least one sound event, and synthesizing stereo sound.

This disclosure provides methods, devices, and systems for audio signal mixing. The present implementations more specifically relate to mixing audio signals from a microphone array by performing fixed beamforming to generate beams, reducing noise on the beams, and mixing the beams to generate a final audio signal for playback. In some aspects, an audio mixing system includes a fixed beamformer to generate beams from audio signals from a microphone array and noise reduction units (NRUs) to reduce a noise component of each audio beam. The system also includes logic to calculate a signal characteristic of each reduced noise audio beam to determine, based on the signal characteristics, the reduced noise audio beams that include a speech component. The logic also generates a gain for each audio beam based on the selection, with the gains used in beam mixing. In some aspects, the NRU includes a neural network noise reduction unit.

The invention provides an audio processing method. The audio processing method comprises the following steps: determining a target sound effect level of a target scene sound effect; according to a preset weight range of each sound effect level in a plurality of sound effect levels and the target sound effect level, determining a target level weight of each sound effect level, the target sound effect level being any one of the plurality of sound effect levels; according to the target hierarchy weight of each sound effect hierarchy, performing audio track mixing on the target sound effect of each sound effect hierarchy to obtain a target scene sound effect of a target scene corresponding to the target sound effect hierarchy; the audio effect can be adjusted by adjusting the weight proportion of different sound effect levels, audio mixing is performed according to the target level weight of each sound effect level, dynamic audio mixing can be realized, immersion and dynamic adaptability can be enhanced, and user experience can be improved.

The invention discloses a method and a device for processing multi-sound-source audio, and electronic equipment. The method comprises the following steps: acquiring audio data input by the multi-sound-source equipment through an input interface by an audio input module; the audio component and the data link module process the audio data according to the audio task, and send the processed audio data to the sound effect pre-processing module; the sound effect pre-processing module carries out sound effect adjustment according to the sound effect configuration information of the audio data to generate a sound effect pre-processing result, and sends the sound effect pre-processing result to the sound mixing module; a sound mixing module carries out sound mixing on the sound effect pre-processing result to generate a sound mixing result, and sends the sound mixing result to a sound effect post-processing module; and the sound effect post-processing module dynamically adjusts the sound mixing result to generate a sound effect post-processing result, and outputs the sound effect post-processing result to the audio output module through the output interface. And audio data is processed in a low-delay manner by using audio task fine-grained componentization, and flexible configuration and sound effect management of multi-sound-source input and output are supported.

The invention provides a broadcast audio processing method and device for a broadcast intelligent audio mixing technology, and the method comprises the steps: collecting an audio signal through an audio interface, carrying out the preprocessing of the audio signal to obtain a first audio signal, and carrying out the feature extraction of the first audio signal; integrating the collected feature vectors to generate a vector matrix as the input of a preset parameter optimization prediction model; receiving a first parameter group output by the parameter optimization prediction model, processing the audio signal through an anti-distortion processing chain, and avoiding distortion caused by audio dynamic problems due to amplitude limiting processing; the audio output quality is monitored in real time, the output quality and the preset ideal audio quality are evaluated, the first parameter group is dynamically adjusted and fed back to the parameter optimization prediction model, and the problems that manual adjustment of multiple paths of audio tracks is time-consuming and prone to errors are solved.

The present application relates to the technical field of digital professional audio equipment, and in particular to a mixing console tuning management method, system, device and storage medium. The present application first obtains the audio signal and the user's mixing selection information to determine the overall mixing numerical information; then, by analyzing the propagation distance, targeted sound parameters are generated for each area, which can better simulate the real sound field, make the sound effect in different areas more natural and realistic, and achieve more refined spatial control; during the actual playback process, the method obtains the real-time sound field monitoring data of each partition, and compares it with the preset ideal sound field model. If the deviation exceeds the threshold range, the sound field optimization is triggered, and the mixing parameters of each audio track are adaptively adjusted according to the optimization feedback, so that the final sound field effect is close to the ideal state, thereby comprehensively improving the efficiency and quality of automatic tuning.

A method for determining a decoding L×K matrix for decoding incoming L-dimensional channel audio into outgoing K-dimensional channel audio where L≥2 and K≥1 is provided. The method comprising: determining a panning control parameter p and a sample component d that minimizes a first difference metric between an L-dimensional input sample x and an estimation of the input sample xest=d a, where a=A(p) and where A(p) is a first pre-set mapping function that returns an L-dimensional panning vector a for a given panning control parameter p; generating a K-dimensional raw output sample yraw=d s, where s=S(p) and where S(p) is a second pre-set mapping function that returns a K-dimensional panning vector s for a given panning control parameter p, and; determining the decoding L×K matrix M by solving an optimization problem that minimizes a second difference metric between the K-dimensional raw output sample yraw and the decoded input sample x M. A method for decoding incoming L-dimensional channel audio into outgoing K-dimensional channel audio using the decoding L×K matrix is also provided.

The invention provides a method, a device and a system for audio signal mixing. The present implementation more particularly relates to mixing audio signals from a microphone array to generate beams by performing fixed beamforming, reducing noise on the beams, and mixing the beams to generate a final audio signal for playback. In some aspects, an audio mixing system includes a fixed beamformer to generate beams from audio signals from a microphone array and a noise reduction unit (NRU) to reduce noise components of each audio beam. The system also includes logic to compute a signal characteristic of each de-noised audio beam to determine a de-noised audio beam including a speech component based on the signal characteristic. The logic also generates a gain for each audio beam based on the selection, wherein the gains are used in beam mixing. In some aspects, the NRU includes a neural network noise reduction unit.

The invention discloses an SIP telephone audio redirection output method and system based on an SIP voice server, and the method comprises the steps: S1, enabling an SIP sound amplification soft terminal in the SIP voice server to be connected to the SIP voice server in a registration mode, and carrying out the qualification authentication and parameter configuration, so as to obtain the audio channel data of a specified SIP telephone on the SIP voice service; s2, the SIP voice service redirects appointed SIP telephone audio data to an appointed Domain Socket of the SIP loudspeaking soft terminal through an internal protocol; s3, the SIP sound amplification soft terminal obtains audio data of a plurality of SIP telephones, and efficient coding, decoding and sound mixing processing are carried out; and S4, the SIP sound amplification soft terminal sends the audio data to the FPGA module through DMA for signal conversion, and then the audio data is directionally output through a loudspeaker. The invention discloses an SIP (Session Initiation Protocol) telephone audio redirection output method and an SIP telephone audio redirection output system based on an SIP voice server, which realize efficient redirection and output of one or more SIP telephone audios, remarkably improve the quality of voice communication and user experience, and provide an efficient and reliable audio solution while reducing the cost.

The invention provides a conference audio processing method and system based on multichannel broadcast and a medium. The method comprises the steps of obtaining conference reservation information, determining corresponding conference spokesman associated data and a semantic feature library according to the conference reservation information, obtaining a real-time conference single-path audio stream, and performing voice signal processing in combination with the semantic feature library and the conference spokesman associated data to obtain an optimized conference single-path audio stream. The method comprises the following steps: acquiring a speaking video and a speaking document image of a conference spokesman, performing spokesman separation and dynamic sound mixing processing according to the speaking video and the speaking document image in combination with an optimized conference single-path audio stream to obtain a multi-channel mixed audio stream, and performing multi-layer sound field reconstruction processing according to the multi-channel mixed audio stream to obtain an adaptive audio; according to the invention, through dynamic beam adjustment, spokesman separation, dynamic sound mixing processing and multi-layer sound field reconstruction, conference audio intelligent processing based on multi-channel broadcast is realized.

An audio processing method and an electronic device are provided. In a dual-view recording mode, the electronic device perform filtering, azimuth virtualization, and remixing on collected audio based on focal lengths of pictures in two display regions, relative positions of the two display regions, and values of areas of the two display regions, so that audio-picture presentation synchronization of the picture and the sound is achieved, and a user has synchronized three-dimensional experience in terms of hearing and vision.