16 results about "Speech enhancement algorithm" patented technology

The invention provides a multi-modal speech enhancement algorithm based on semi-supervised learning pre-training. The multi-modal speech enhancement algorithm comprises the following steps: S1, taking an extracted noisy speech feature vector # imgabs0 # as the input of a noisy speech GCRBM; s2, de-noising the noisy voice # imgabs 1 # by using a traditional spectral subtraction method to obtain a de-noised voice # imgabs 2 #; s3, taking the estimated de-noised speech feature vector # imgabs3 # as the input of the pure speech GCRBM; s4, taking the extracted video signal feature vector # imgabs4 # as the input of the video GCRBM; s5, optimizing parameters of the multi-mode speech enhancement network; s6, during testing, noise-containing speech features are input to the multi-mode speech enhancement network; according to the method, the voice features and the mouth features are extracted and input into the multi-mode voice enhancement network, and after parameter optimization of the multi-mode voice enhancement network is carried out, the noisy voice features are input into the multi-mode voice enhancement network to obtain the enhanced voice signals, so that the voice annotation cost is reduced, and the intelligibility of voice enhancement is improved.

The invention discloses a distributed reconnaissance tool, relates to the technical field of voice interception, and aims to solve the problems of limited multi-node access capability, poor anti-interference performance and insufficient multi-target interception continuity of an existing reconnaissance system. The tool comprises a micro audio node network and a base station unit, a recording module, a wireless transmission module and an encryption module are arranged in the miniature audio node, local recording and remote transmission are supported, and the miniature audio node has the characteristics of water resistance and low power consumption; the base station unit adopts an orthogonal graph division multiple access technology, at most 32 micro audio nodes can be accessed, four paths of voice can be monitored in real time, the audio quality is optimized through a spectral subtraction voice enhancement algorithm, and WIFI / Ethernet connection and time synchronization and combination of multi-node recording files are also supported. According to the invention, approaching interception and multi-target centralized deployment and control of the moving target are realized, the stability, expansibility and practicability of the interception system are improved, and the method is suitable for scenes such as public security technical investigation and the like.

The invention provides a speech enhancement algorithm evaluation method and device, electronic equipment, a chip and a storage medium, and the method comprises the steps: carrying out the noise reduction of a noisy speech through employing a speech enhancement algorithm, and obtaining a noise-reduced speech; the noise reduction processing process is evaluated through a plurality of evaluation indexes, a plurality of evaluation results are obtained, the evaluation results correspond to the evaluation indexes, and the evaluation indexes comprise one or more of parameter quantity, floating point operand FLOPs, perceptual speech quality PESQ and short-term objective intelligibility STOI; the multiple evaluation indexes are converted to the same order of magnitude, multiple conversion results are obtained, and the conversion results correspond to the evaluation results; according to the method, the multiple objective indexes are used for evaluating the speech enhancement algorithm, the multiple objective indexes are converted to the same order of magnitude for fusion processing, and the comprehensive and objective evaluation result is obtained.

The invention discloses a single-microphone long-distance pickup method and device, and the method comprises the steps: determining a microphone bias voltage according to the microphone type, circuit and characteristics of an employed single microphone, enabling the single microphone to directly face a sound source for pickup, and obtaining a to-be-processed audio signal; automatically adjusting the overall level of the audio signal to be processed by adopting an automatic level control algorithm; converting the audio signal to be processed after the level adjustment into a digital signal; a voice enhancement algorithm and an automatic gain control algorithm are adopted to carry out noise reduction, voice enhancement and gain adjustment processing on the digital signal in sequence, that is, a method of combining an analog domain and a digital domain is used to increase the pickup distance; the quality of audio signals acquired by a single microphone is optimized by using an automatic level control (ALC) technology.

The present application provides a method and apparatus for speech quality evaluation, which relates to the field of artificial intelligence. The method includes: enhancing the speech signal collected by a device through a speech enhancement algorithm to obtain enhanced speech; inputting the enhanced speech into a speech quality evaluation model to obtain loss information of the enhanced speech, where the speech quality evaluation model includes at least two acoustic models, and the above loss information is determined based on the decoding results of the at least two acoustic models' decoders for the enhanced speech as the ground truth; and determining whether the quality of the enhanced speech meets the standard based on the loss information. The above speech quality evaluation method and apparatus can be decoupled from a specific speech enhancement algorithm and a specific ASR backend, can reflect the influence of the speech enhancement algorithm on the ASR backend, and can be flexibly applied to a variety of different scenarios.

The invention provides a tinnitus patient intelligent social assistance system and method based on multi-modal fusion, and relates to the technical field of medical assistance, and the method comprises the steps: collecting environment sound signals, voice signals and image data, and carrying out the primary processing; a personalized adaptive noise reduction algorithm and a voice enhancement algorithm are formulated, and then an enhanced voice signal is obtained; constructing a lip language recognition model, and performing conversion correction processing on the enhanced voice signal; constructing a tinnitus feature database, and screening tinnitus masking sound effects from the tinnitus feature database according to the individual tinnitus parameters; obtaining an association relationship between tinnitus and environmental factors during social contact of the tinnitus patient, and providing personalized prevention suggestions according to the association relationship and the tinnitus masking sound effect; and uploading the corresponding enhanced voice signal and the preliminarily processed image data to a cloud server so as to optimize a personalized adaptive noise reduction algorithm and a voice enhancement algorithm.

A method for enhancement of speech intelligibility in a device arranged for a near-end side a communication with a far-end device. The method involves calculating a measure of speech intelligibility at the near-end side based on a near-end audio input and a far-end audio input. Then, based on the calculated measure of speech intelligibility optimizing parameters of a predetermined speech enhancement algorithm, where a predetermined speech intelligibility target, and an additional target are taken into account to generate an optimized speech enhancement algorithm. Next, processing the far-end audio input according to the optimized speech enhancement algorithm, and generating a near-end audio output accordingly. The algorithm can adapt to changing noise conditions and be optimized for both speech intelligibility and another target. This can be used to minimize delay, electric power consumption and audio quality while satisfying the speech intelligibility target. The optimization can be based on a closed-form solution.

The invention discloses a marine miniaturized network cabin voice recording method and a marine miniaturized network cabin voice recording device, which relate to the technical field of shipborne recording equipment, can realize a strong-pointing pickup function, effectively eliminate environmental noise, enhance voice and intelligently judge a voice instruction of a specific person (such as a commander) needing to be extracted through voiceprint recognition. A directional pickup function is realized by using a beam forming technology based on a microphone array, and enhancement is performed aiming at the position of a specific commander in a cabin, so that interference caused by environmental noise can be effectively reduced. And detecting whether the picked sound contains human voice by using a human voice detection algorithm, if the picked sound contains the human voice, further enhancing the picked sound by using a voice enhancement algorithm, otherwise, if the picked sound does not contain the human voice, estimating the environmental noise. In addition, a commander instruction extraction function based on voiceprint recognition is also realized, and the command voice instruction of the commander is automatically extracted according to pre-configured commander identity information.

The invention provides a multi-channel speech enhancement method for a teleconference, which comprises the following steps of: firstly, scanning the amplitudes of plane waves in different directions, finding the direction of a speaker, collecting a sound signal, then adjusting the phase and gain of the sound signal based on a speech enhancement algorithm, enhancing the signal, and finally playing at a remote end. Compared with the prior art, the method has the beneficial effects that sound signals in the direction of a speaker are enhanced, noise signals in other directions are inhibited, the influence of reverberation is reduced, and the call quality is improved; a frequency-independent beam former and a traditional delay summation beam former are combined, the respective defects of the two beam formers are overcome, and the beam formers have high signal-to-noise ratios no matter low-frequency signals or high-frequency signals.

The application discloses a speech enhancement method, comprising: acquiring audio information collected by a robot; generating a speech presence probability of speech included in the audio information according to the audio information; and enhancing the audio information according to the speech presence probability to generate enhanced speech. According to the embodiment of the application, the speech presence probability of speech is generated according to the audio information, and the audio information can be enhanced according to the speech presence probability when the speech is enhanced, so that the enhancement effect is improved. In the embodiment of the application, the precision of the speech enhancement algorithm can be improved through the speech presence probability, so that the speech enhancement effect is realized.

The application discloses a single microphone long-distance sound pickup method and device, which comprises the following steps: determining a microphone bias voltage according to the microphone type, circuit and characteristics of the single microphone to be used, directly picking up sound from the sound source by the single microphone to obtain a to-be-processed audio signal; automatically adjusting the overall level of the to-be-processed audio signal by using an automatic level control algorithm; converting the to-be-processed audio signal after the level adjustment into a digital signal; sequentially performing noise reduction, speech enhancement and gain adjustment processing on the digital signal by using a speech enhancement algorithm and an automatic gain control algorithm, that is, using a method combining an analog domain and a digital domain to improve the sound pickup distance; and using an automatic level control (ALC) technology to optimize the quality of the audio signal collected by the single microphone.

To provide a method for processing audio signals, as well as a corresponding device, a computer program, and computer-readable storage media.SOLUTION: A method 1000 includes the steps of: applying emphasis to a first component of an audio signal and / or applying suppression to a second component of the audio signal with respect to the first component S1010; and modifying an output of the step S1010 by applying a deep learning based model to the output to remove artifacts and / or distortions introduced into the audio signal in the step S1010, and perceptually improving the first component of the audio signal S1020.SELECTED DRAWING: Figure 10

This paper provides an improved, high-performance noise reduction algorithm based on the self-attention mechanism. This algorithm leverages the powerful mapping capabilities of deep neural networks to reduce the computational complexity of the self-attention mechanism. Combined with a spectral masking strategy, it achieves a high-performance noise reduction algorithm for complex noisy environments. Compared to other speech enhancement algorithms, this algorithm has lower computational complexity, effectively captures the time-frequency characteristics of the input speech, and achieves better noise reduction in low signal-to-noise ratio environments.

The present invention proposes a speech enhancement algorithm based on improved phase spectrum compensation and fully convolutional neural network. The feature data of noisy speech and clean speech are obtained by preprocessing the speech data in the training set. Combining the feature data of noisy speech and clean speech, the compensation factor calculation formula of the frame signal-to-noise ratio optimized phase compensation algorithm is introduced, and then the improved formula is used to calculate the phase compensation factor. A fully convolutional neural network model is built, and the phase compensation factor combined with the logarithmic power spectrum of clean speech is used as the training target of the fully convolutional neural network for network model training. The test speech is input into the trained model to obtain the estimated value of the logarithmic power spectrum and the phase compensation function. The amplitude spectrum and phase spectrum of the speech signal are reconstructed respectively using the estimated value of the logarithmic power spectrum and the phase compensation function to obtain the final enhanced speech. The present invention improves the noise cancellation ability of the algorithm while better ensuring the speech intelligibility, thus improving the overall effect of speech enhancement.

The application discloses a double-branch speech enhancement algorithm based on a structured state space sequence model, which comprises the following steps: obtaining amplitude spectrum and complex spectrum features of noisy speech, and inputting the features into an amplitude rough estimation branch and a complex refinement estimation branch respectively to obtain real and imaginary components of the rough estimated speech and the refined speech; introducing an interaction module to realize the flow of the amplitude spectrum and the complex spectrum features between the two branches; superimposing the real and imaginary components of the rough estimated speech and the refined speech to reconstruct a target signal complex spectrum; and evaluating the performance of the double-branch enhancement algorithm based on the structured state space sequence model. The amplitude spectrum and the complex spectrum are estimated simultaneously, and the interaction module is introduced to promote information exchange, so that the features learned from one branch can supplement the missing information of the other branch; and a diagonalized state space model is used to model the speech feature sequence, so that the parameter quantity of the model is reduced, and the algorithm performance is improved.

The invention discloses a method for improving voice definition of a bone conduction earphone. The bone conduction earphone comprises a bone conduction transducer, a signal processing unit, a voice definition improving module and an earphone body, the signal processing unit is electrically connected with the bone conduction transducer and the voice definition improving module; the earphone body is provided with at least one bone conduction transducer; the voice definition improving module comprises at least one filter and a voice enhancing algorithm processor; the filter is used for primarily processing the signal received by the bone conduction transducer; and the speech enhancement algorithm processor is used for further processing the filtered signal so as to improve the speech definition.