Over the past few years, audio enhancement methods (e.g., for recorded human speech) based on deep learning have greatly surpassed traditional methods (e.g., due to techniques such as spectral subtraction and spectral estimation). Audio enhancement methods may be used in a variety of applications. For example, a teleconferencing system may be used in a noisy and reverberant environment, so speech enhancement techniques may be needed to ensure clear communication.
By using multiple microphones for a teleconference, spatial aspects of sound may be captured, such as the locations of speakers or other sound sources. It may be useful to pursue high speech enhancement performance while at the same time preserving spatial cues because spatial cue information may help a listener to determine who is speaking during a teleconference. However, the large maintenance and training cost to implement a stereo-specific speech enhancement model (e.g., a deep neural network (DNN) model) can become prohibitive. Furthermore, applying a single-channel noise suppressor independently to each microphone signal may not preserve the spatial location of the speech and may also result in audible artifacts, particularly if the spatial properties of the target speech and the interfering noise are different.
While embodiments are described herein by way of example for several embodiments and illustrative drawings, those skilled in the art will recognize that embodiments are not limited to the embodiments or drawings described. It should be understood, that the drawings and detailed description thereto are not intended to limit embodiments to the particular form disclosed, but on the contrary, the intention is to cover all modifications, equivalents and alternatives falling within the spirit and scope as described by the appended claims. The headings used herein are for organizational purposes only and are not meant to be used to limit the scope of the description or the claims. As used throughout this application, the word “may” is used in a permissive sense (i.e., meaning having the potential to), rather than the mandatory sense (i.e., meaning must). Similarly, the words “include,” “including,” and “includes” mean including, but not limited to.
It will also be understood that, although the terms first, second, etc. may be used herein to describe various elements, these elements should not be limited by these terms. These terms are only used to distinguish one element from another. For example, a first contact could be termed a second contact, and, similarly, a second contact could be termed a first contact, without departing from the scope of the present invention. The first contact and the second contact are both contacts, but they are not the same contact.
Various techniques for real-time low-complexity stereo speech enhancement with spatial cue preservation are described herein. With the ubiquitous presence of real-time audio communication systems, there has been a significant interest in speech enhancement algorithms that operate in real-time with low complexity. Users (e.g., a target speaker) of these communication systems may find themselves in the presence of competing background sounds (e.g., noise from various sources). In various embodiments, any number of microphones may be used during a teleconference, resulting in multiple input signals. For example, two microphones may be used to provide a stereo input.
Implementing a stereo-specific speech enhancement model (e.g., a highly complex deep neural network (DNN) or other type of model) can become prohibitive. Instead, various embodiments may use a low-complexity model and/or algorithm, such as the example “PercepNet” model discussed herein, to deliver high-quality speech enhancement in real-time. As discussed in more detail below, in order to leverage such a model, the stereo input is downmixed to a monaural signal (while also estimating spatial cues, e.g., a steering vector and/or interaural phase/level differences between the different inputs), processed by the model, and then upmixed to produce an enhanced stereo signal (taking into account the spatial cues).
In various embodiments, machine learning (ML) model-based audio enhancement techniques that provide a perceptually motivated approach to real-time, low-complexity target speaker enhancement, such as PercepNet, may utilize techniques for real-time target speaker audio enhancement for a monaural signal (e.g., after estimating spatial cues and downmixing from the stereo signal). Real-time target speaker audio enhancement techniques may perform conditioning on the target speaker's voice and/or other audio enhancement techniques. This enables the machine learning model for audio enhancement to distinctly identify and enhance the target speaker's utterance while suppressing all the other interferences, even in the presence of multiple talkers or other speech-like sounds. As discussed below, the enhanced monaural signal is then upmixed based on the estimated spatial cues for the target speaker in order to generate the enhanced stereo output signal. In some embodiments, the above process may be performed for any number of different target speakers.
In some embodiments, it may be assumed that there are two microphone input signals (note that in various embodiments, any number of microphone inputs may be used). The m-th microphone input signal xm,t (t is the time-index) may be modeled in the time domain as follows: xm,t=st*hm+nt, where st is the speech source signal, hm is the impulse response between the speech source location and the m-th microphone, and nt is the background noise signal. In embodiments, downmix, monaural speech enhancement based on a model (e.g., PercepNet or other model), and upmix are all carried out in the time-frequency domain. In embodiments, the time-frequency representation of xm,t can be written as follows: xl,k=sl,kak+nl,k, where l is the frame-index, k is the frequency index, xl,k=[xT1,l,k xT2,l,k]T, xm,l,k is the time-frequency representation of the time-domain signal xm,t, sl,k and nl,k are defined similarly. ak=[a1,k a2,k]T is a steering vector and am,k is the time-frequency representation of hm.
Typically, the objective of speech enhancement is to estimate sl,k from the observed microphone input signal xl,k. In our application, the steering vector ak is also important in that it captures spatial information, so our objective is to determine both sl,k and ak from the microphone input signal xl,k. Additionally, since it is difficult and costly to retrain a DNN model for different numbers and locations of sound sources and microphones, we focus on using a single-channel DNN model pre-trained for monaural speech enhancement.
In embodiments, an example stereo speech enhancement system 100 may include three steps/stages. In the first step, spatial cues for one or more target sources (e.g., speakers) represented in a stereo microphone input signal are estimated and stored as metadata (e.g., interaural phase/level differences between the different inputs for one or more target speakers), and the stereo microphone input signal is converted into a monaural signal (downmix 102). The second step includes the use of an ML model 104 and/or algorithm to enhance the monaural signal (e.g., PercepNet or other ML model). In an embodiment, the algorithm may operate on 10 ms frames with 30 ms of look-ahead, although any other amounts of time may be used in various embodiments. As mentioned above, in order to utilize a model of a single microphone as a pre-trained model (e.g., a DNN model or other pretrained ML model), the input signal for the model is generated by downmixing a stereo microphone input signal into a monaural input signal (spatial information/cues are preserved). A stereo signal may be generated from the enhanced monaural output signal of the model, by applying the preserved spatial information (e.g., applying the interaural phase and level differences to the monaural signal to generate two different output signals). Therefore, the third step includes converting the output signal of the model into a stereo signal (upmix 106).
In various embodiments, the stereo speech enhancement system 100 may be implemented as part of various network-based systems or services or stand-alone systems that receive audio data (e.g., stereo speech audio, which may include target speaker audio and various background audio from a first input signal from a left microphone and a second input signal from a right microphone) and provide as output enhanced audio data (e.g., enhanced stereo speech audio, which may include enhanced target speaker audio and various background audio of a first output signal and a second output signal). For example, a stereo speech enhancement system 100 may be implemented “service-side,” as illustrated in
In embodiments, the stereo speech enhancement system may receive, via an interface for the stereo speech enhancement system, a stereo input signal that includes a first input signal and a second input signal (e.g., a left and right input signal respectively corresponding to a left and right microphone or a right and a left input signal respectively corresponding to a right and left microphone). The first input signal may include speech data corresponding to a target speaker (e.g., generated by an audio sensor/microphone that senses the target speaker's voice) and the second input signal may include other speech data corresponding to the target speaker (e.g., generated by another audio sensor/microphone at a different location that senses the target speaker's voice).
The stereo speech enhancement system may then downmix the stereo input signal to generate a monaural signal. Before downmixing the signal, spatial cues for the target speaker are first estimated and preserved (e.g., stored as metadata) using any suitable technique (e.g., storing the spatial cues to a temporary storage location in memory or other type of data store/database). The stereo speech enhancement system then applies a ML model to the monaural signal to generate an enhanced monaural signal. In embodiments, noise in the monaural signal is reduced or removed from the monaural signal by the ML model in order to generate the enhanced monaural signal.
The stereo speech enhancement system may then upmix the enhanced monaural signal based at least on the spatial cues for the target speaker to generate an enhanced stereo output signal. The enhanced stereo output signal includes a first output signal and a second output signal. The first output signal includes enhanced speech data corresponding to the target speaker and the second output signal comprises other enhanced speech data corresponding to the target speaker. The stereo speech enhancement system may then send, via the interface of the stereo speech enhancement system, the enhanced stereo output signal to a destination. Note than in some embodiments, only a partially processed stereo signal may be provided, depending on the capabilities or needs of the target destination. For example, if the target destination only supports monaural playback (e.g., only has one speaker), then the enhanced monaural signal may be sent to the destination (in some cases, the estimated spatial cues may also be sent). In some embodiments, the monaural signal and/or the estimated spatial cues may be sent to a destination (e.g., before any processing may a model).
As described herein, various techniques may be used by the stereo speech enhancement system to perform the upmixing and/or the downmixing stages. For example, downmixing the stereo input signal may include applying beamforming to the stereo input signal (e.g., delay-and-sum beamforming). In some embodiments, to estimate the spatial cues (e.g., spatial information), the stereo speech enhancement system may estimate steering information (e.g., an estimated steering vector) of the target speaker. The system may apply the delay-and-sum beamforming to the stereo input signal using the estimated steering information (e.g., the estimated steering vector) to generate the monaural signal.
In some embodiments, upmixing the enhanced monaural signal may include applying steering information of the target speaker to the enhanced monaural signal (e.g., multiplying an estimated steering vector of the target speaker with the enhanced monaural signal). In embodiments, the stereo speech enhancement system may estimate a steering vector of the target speaker using principal component analysis of an estimated spatial covariance matrix. In some embodiments, the stereo input signal is captured along with corresponding video data, and the video data may be provided to a same destination as the enhanced stereo output signal.
This specification includes a general description of a provider network that implements multiple different services (
Downmix
Using traditional techniques, downmixing may be performed by averaging a stereo signal. However, averaging may cause cancellation of a speech signal and may lead to performance degradation. Instead of averaging a stereo signal, conversion of a stereo input signal into a monaural signal is done by using a beamforming technique, in embodiments. Downmix may be done by using delay and sum beamforming (DSBF) for the purpose of avoiding signal cancellation related to adaptive beam-forming techniques. Let blk be the estimated steering vector. The downmixed monaural signal dlk is obtained as follows: dlk=bHl,k xl,k, where H is the Hermitian transpose operator of a matrix/vector. To perform DSBF, the steering vector of the speech source blk is estimated. blk is estimated by using principal component analysis (PCA) with an estimated spatial covariance (SCM).
Spatial Covariance Matrix Estimation
To perform real-time speech enhancement, a SCM Rl,k is updated in an online manner as follows: Rlk=γlkRl−1,k+(1−γlk)xlkxlkH, and γlk=1−lk (1−α), where α is the forgetting factor. lk is a time-frequency mask which controls updating of the SCM depending on presence of a speech source at each time-frequency bin. When a speech source is present, lk is close to 1. On the other hand, when a speech source is absent, lk is close to 0. The time-frequency mask lk may be estimated in two ways: time averaging of microphone input signal (TAM) or time-frequency masking based on speech enhancement output (TFMSE).
Time Averaging of Microphone Input Signal (TAM)
When there is no information about the presence of the speech source at each time-frequency bin, the best way is to assume that there is always a speech source. In this case, the time-frequency mask should be determined as follows: lk=1. This is corresponding to time-averaging of microphone input signal for SCM estimation. When noise signal is uncorrelated between channels and noise level is relatively lower than speech signal level, time-averaging works well. However, SCM estimation accuracy degrades when the noise level is relatively bigger than speech signal level or noise signal is correlated between channels.
Time-Frequency Masking Based on Speech Enhancement Output (TFMSE)
To avoid degradation of performance in SCM estimation, it may be important to extract time-frequency bins in which speech sources are dominant. Application of monaural speech enhancement result for estimation of time-frequency masks may be performed as follows: lk=∥clk∥/∥xlk∥, where clk is an enhanced speech signal. clk can be generated by multiplying a steering vector with a monaural speech enhance-ment result. Let slk be the output monaural signal after speech enhancement. So as to compare ŝlk with the microphone in-put signal, a temporal upmix is done, and we can obtain stereo noiseless signal as follows: ĉlk=ŝlkâlk, where âlk is the estimated steering vector of the speech signal. Block diagram of this approach is shown in
Upmix
Upmix converts a monaural signal into a stereo signal. Upmix may be performed in two ways: steering vector multiplication or a common gain-based method.
Steering Vector Multiplication
Steering vector multiplication converts a monaural signal into a stereo signal by multiplying an estimated steering vector with the monaural signal. In embodiments, the steering vector may be estimated for the upmix stage in the same or similar way as it is estimated for the downmix stage. Let elk be the estimated steering vector for upmix. The upmixed signal may be obtained as follows: flk=dlkelk, where f is the output stereo signal. Similar to downmix, the steering vector is estimated via PCA of the estimated SCM. We can utilize TAM and TFMSE for SCM estimation.
Common Gain-Based Method
Sharing time-frequency gain between channels is known as an effective approach for spatial-cue preservation [25]. Our common gain based method shares 32 band gains estimated in PercepNet between channels. The common gain has been typically applied for a noisy microphone input signal. We call this method common1. However, common1 lacks of utilization of beamforming capability. Another approach is to apply the common band gain for the enhanced speech signal, i.e., clk in Eq. 8. We call this method common2. In common2, upmix can be interpreted as a hybrid combination of steering vector multiplication and a common band gain multiplication.
In various embodiments, a hybrid framework combines model-based monaural speech enhancement (e.g., DNN model) and array signal processing as a real time stereo speech enhancement technique that preserves spatial cues. Both spatial cue preservation performance and speech enhancement performance can be improved by using beamforming based downmix and steering vector multiplication based upmix. In embodiments, time-frequency masking based steering vector estimation and/or a common gain approach for a beamformer output may be quite effective. In some embodiments, one DNN inference per frame may be sufficient using techniques discussed herein. Using embodiments discussed herein, there may be a significant advantage in terms of computational cost over traditional techniques, especially when the number of microphones is large. Furthermore, embodiments may be easily extended to more general multichannel speech enhancement algorithms without DNN retraining.
Provider network 200 may be a private or closed system or may be set up by an entity such as a company or a public sector organization to provide one or more services (such as various types of cloud-based storage) accessible via the Internet and/or other networks to clients 250, in one embodiment. Provider network 200 may be implemented in a single location or may include numerous data centers hosting various resource pools, such as collections of physical and/or virtualized computer servers, storage devices, networking equipment and the like (e.g., computing system 1200 described below with regard to
In various embodiments, the components illustrated in
Audio-transmission service 210 may implement interface 211 to allow clients (e.g., client(s) 250 or clients implemented internally within provider network 200, such as a client application hosted on another provider network service like an event driven code execution service or virtual compute service) to send audio data (e.g., stereo speech input signals) for enhancement, storage, and/or transmission. In at least some embodiments, audio-transmission service 210 may also support the transmission of video data along with the corresponding audio data and thus may be an audio/video transmission service, which may perform the various techniques discussed above with regard to
Audio-transmission service 210 may implement a control plane 212 to perform various control operations to implement the features of audio-transmission service 210. For example, control plane 212 may monitor the health and performance of requests at different components audio-transmission 213 and audio enhancement 215 (e.g., the health or performance of various nodes implementing these features of audio-transmission service 210). If a node fails, a request fails, or other interruption occurs, control plane 212 may be able to restart a job to complete a request (e.g., instead of sending a failure response to the client). Control plane 212 may, in some embodiments, may arbitrate, balance, select, or dispatch requests to different node(s) in various embodiments. For example, control plane 212 may receive requests interface 211 which may be a programmatic interface, and identify an available node to begin work on the request.
Audio-transmission service 210 may implement audio-transmission 213, which may facilitate audio communications (e.g., for audio-only, video, or other speech communications), speech commands or speech recordings, or various other audio transmissions, as discussed in the examples below with regard to
Data storage service(s) 230 may implement different types of data stores for storing, accessing, and managing data on behalf of clients 250 as a network-based service that enables clients 250 to operate a data storage system in a cloud or network computing environment. Data storage service(s) 230 may also include various kinds relational or non-relational databases, in some embodiments. Data storage service(s) 230 may include object or file data stores for putting, updating, and getting data objects or files, in some embodiments. Data storage service(s) 230 may be accessed via programmatic interfaces (e.g., APIs) or graphical user interfaces. Enhanced audio 232 may be put and/or retrieved from data storage service(s) 230 via an interface for data storage services 230, in some embodiments, as discussed below with regard to
Generally speaking, clients 250 may encompass any type of client that can submit network-based requests to provider network 200 via network 260, including requests for audio-transmission service 210 (e.g., a request to enhance, transmit, and/or store audio data). For example, a given client 250 may include a suitable version of a web browser, or may include a plug-in module or other type of code module that can execute as an extension to or within an execution environment provided by a web browser. Alternatively, a client 250 may encompass an application (or user interface thereof), a media application, an office application or any other application that may make use of audio-transmission service 210 (or other provider network 200 services) to implement various applications. In some embodiments, such an application may include sufficient protocol support (e.g., for a suitable version of Hypertext Transfer Protocol (HTTP)) for generating and processing network-based services requests without necessarily implementing full browser support for all types of network-based data. That is, client 250 may be an application that can interact directly with provider network 200. In some embodiments, client 250 may generate network-based services requests according to a Representational State Transfer (REST)-style network-based services architecture, a document or message-based network-based services architecture, or another suitable network-based services architecture.
In some embodiments, a client 250 may provide access to provider network 200 to other applications in a manner that is transparent to those applications. Clients 250 may convey network-based services requests (e.g., requests to interact with services like audio-transmission service 210) via network 260, in one embodiment. In various embodiments, network 260 may encompass any suitable combination of networking hardware and protocols necessary to establish network-based-based communications between clients 250 and provider network 200. For example, network 260 may generally encompass the various telecommunications networks and service providers that collectively implement the Internet. Network 260 may also include private networks such as local area networks (LANs) or wide area networks (WANs) as well as public or private wireless networks, in one embodiment. For example, both a given client 250 and provider network 200 may be respectively provisioned within enterprises having their own internal networks. In such an embodiment, network 260 may include the hardware (e.g., modems, routers, switches, load balancers, proxy servers, etc.) and software (e.g., protocol stacks, accounting software, firewall/security software, etc.) necessary to establish a networking link between given client 250 and the Internet as well as between the Internet and provider network 200. It is noted that in some embodiments, clients 250 may communicate with provider network 200 using a private network rather than the public Internet.
Sensor(s) 252, such as microphones, may, in various embodiments, collect, capture, and/or report various kinds of audio data, (or audio data as part of other captured data like video data). Sensor(s) 252 may be implemented as part of devices, such as various mobile or other communication and/or playback devices, such as microphones embedded in “smart-speaker” or other voice command-enabled devices. In some embodiments, some or all of audio enhancement techniques may be implemented as part of devices that include sensors 252 before transmission of enhanced audio to audio-transmission service 210, as discussed below with regard to
As discussed above, different interactions between sensors that capture audio data and services of a provider network 200 may invoke audio enhancement, in some embodiments.
In
Audio-transmission service 210 may process captured audio data 312 through audio enhancement 215 (e.g., stereo speech enhancement), in various embodiments. For example, an audio enhancement systems like those discussed below with regard to
Audio enhancement systems may also be implemented separately from audio-transmission service 210, in some embodiments. For example, as illustrated in
Device with audio sensor 330 may then send the capture/enhanced audio data 334 to audio-transmission service 210 for transmission (e.g., via interface 211), in some embodiments. Audio transmission 213 may receive the enhanced audio data 334, identify a destination for the enhanced audio, such as audio playback device 340, and send the enhanced audio data 336 to audio playback device 340, in some embodiments.
As mentioned above, in various embodiments, any portions of the audio enhancement process may be performed at the local client network (e.g., by the device with audio sensors 330), and remaining portions of the audio enhancement process may be performed by the provider network. For example, the downmix stage may be performed by a local/client device at a client network, and then the monaural signal (along with the estimated spatial cues) can be transmitted to the provider network in order to perform the remaining processing at the provider network (e.g., the model processing and the upmix processing) to generate the enhanced stereo output. As another example, the model processing of the monaural signal may be performed by a local/client device at a client network, and the enhanced monaural signal may then be transmitted to the provider network (along with the estimated spatial cues) in order to perform the remaining processing at the provider network (e.g., the upmix processing) to generate the enhanced stereo output.
In some embodiments, audio may be stored for later retrieval and/or processing. As illustrated in
The upmix stage 408 performs additional processing on the output of the ML model as well as the input signals. As shown, steering vector multiplication 410 is performed on results of the ML model. Buffering 412 (30 ms delay) receives the left and right input signal and provides the input signals to TFM estimation 414 as well as spatial covariance matrix estimation 416. TFM estimation 414 is performed based on the input signals and the output of steering vector multiplication 410. Spatial covariance matrix estimation 416 is performed based on the input signals and the output of TFM estimation 414. Steering vector estimation 418 is performed based on the output of spatial covariance matrix estimation 416. The steering vector estimation 418 sends its output to the additional steering vector multiplication 420 as well as to the delay-and-sum beamforming 404 of the downmix stage. The steering vector multiplication 420 is performed based on output from the ML model 406 and the steering vector estimation 418 to produce the stereo output.
The output of each ML model 504 is provided for TFM estimation 506 (ML models 504a and 504b may be the same type of model/identical models, in embodiments). Buffering 508 (30 ms delay) also provides the left and right input signals to TFM estimation 506. TFM estimation is performed on the above inputs to provide an output for spatial covariance estimation 510. Spatial covariance estimation 510 is performed based on this output as well as the buffered left and right input signals. The output is provided for steering vector estimation 512. The output of steering vector estimation 512, as well as the left and right input signals, is provided for delay-and-sum beamforming 514 to produce the monaural signal.
As shown, the example algorithm uses DSBF 704 for target source enhancement and DSBF 706 for another source or noise enhancement. The same ML model 708a and 708b (e.g., different ML models that are the same type of model/identical models, in embodiments) is performed for the target source and the other source or noise. As shown, TFM estimation and SCM adaption is also performed 710. In an embodiment, the upmix 702 step may perform steering vector multiplication.
As shown, the example algorithm uses DSBF 804a for target source enhancement and DSBF 804b for another source or noise enhancement. The same ML model 806a and 806b (e.g., different ML models that are the same type of model/identical models, in embodiments) is performed for the target source and the other source or noise. As shown, SCM adaption/steering vector estimation is also performed 808. Mixing 810 is performed on the output of the ML models 806a, 806b to generate the enhanced stereo output.
As shown, the example algorithm uses filtering 904a for target source enhancement and filtering 904b for another source or noise enhancement. The same ML model 906a and 906b (e.g., different ML models that are the same type of model/identical models, in embodiments) is performed for the target source and the other source or noise. As shown, BSE filter adaptation is also performed 908. Mixing 910 is performed on the output of the ML models 906a, 906b to generate the enhanced stereo output.
In various embodiments, any number of microphones may be used during a teleconference, resulting in multiple input signals/channels. The illustrated stereo speech enhancement system 1000 represents an extension of the speech enhancement system described in
Although
As indicated at 1110, a stereo speech enhancement system may receive, via an interface for the stereo speech enhancement system, a stereo input signal (a first input signal and a second input signal). For example, the stereo input signal may be received from two or more audio sensors, as discussed above with regard to
As indicated at 1120, the stereo speech enhancement system estimates spatial cues for a target speaker represented in the stereo input signal and downmixes the stereo input signal to generate a monaural signal. The spatial cues are stored for later use during the upmixing stage (e.g., as metadata/side information). As indicated at 1130, the stereo speech enhancement system may apply a machine learning model to the monaural signal to generate an enhanced monaural signal; noise is reduced or removed to generate the enhanced monaural signal. As indicated at 1140, the stereo speech enhancement system may upmix the enhanced monaural signal based on the estimated spatial cues for the target speaker to generate an enhanced stereo input signal (the stereo signal includes a first output signal and a second output signal). The output signals of the stereo signal may have different interaural phases and/or levels based on the estimated spatial cues, allowing a listener to perceive a location of the speaker and movement of the speaker as the speaker is talking.
As indicated at 1150, the stereo speech enhancement system may provide the enhanced stereo output signal (e.g., stored, transmitted, or otherwise communicated), in some embodiments (e.g., as discussed above with regard to
The methods described herein may in various embodiments be implemented by any combination of hardware and software. For example, in one embodiment, the methods may be implemented on or across one or more computer systems (e.g., a computer system as in
Embodiments of real-time low-complexity speech enhancement with spatial cue preservation as described herein may be executed on one or more computer systems, which may interact with various other devices. One such computer system is illustrated by
In the illustrated embodiment, computer system 1200 includes one or more processors 1210 coupled to a system memory 1220 via an input/output (I/O) interface 1230. Computer system 1200 further includes a network interface 1240 coupled to I/O interface 1230, and one or more input/output devices 1250, such as cursor control device 1260, keyboard 1270, and display(s) 1280. Display(s) 1280 may include standard computer monitor(s) and/or other display systems, technologies or devices. In at least some implementations, the input/output devices 1250 may also include a touch or multi-touch enabled device such as a pad or tablet via which a user enters input via a stylus-type device and/or one or more digits. In some embodiments, it is contemplated that embodiments may be implemented using a single instance of computer system 1200, while in other embodiments multiple such systems, or multiple nodes making up computer system 1200, may host different portions or instances of embodiments. For example, in one embodiment some elements may be implemented via one or more nodes of computer system 1200 that are distinct from those nodes implementing other elements.
In various embodiments, computer system 1200 may be a uniprocessor system including one processor 1210, or a multiprocessor system including several processors 1210 (e.g., two, four, eight, or another suitable number). Processors 1210 may be any suitable processor capable of executing instructions. For example, in various embodiments, processors 1210 may be general-purpose or embedded processors implementing any of a variety of instruction set architectures (ISAs), such as the x86, PowerPC, SPARC, or MIPS ISAs, or any other suitable ISA. In multiprocessor systems, each of processors 1210 may commonly, but not necessarily, implement the same ISA.
In some embodiments, at least one processor 1210 may be a graphics processing unit. A graphics processing unit or GPU may be considered a dedicated graphics-rendering device for a personal computer, workstation, game console or other computing or electronic device. Modern GPUs may be very efficient at manipulating and displaying computer graphics, and their highly parallel structure may make them more effective than typical CPUs for a range of complex graphical algorithms. For example, a graphics processor may implement a number of graphics primitive operations in a way that makes executing them much faster than drawing directly to the screen with a host central processing unit (CPU). In various embodiments, graphics rendering may, at least in part, be implemented by program instructions that execute on one of, or parallel execution on two or more of, such GPUs. The GPU(s) may implement one or more application programmer interfaces (APIs) that permit programmers to invoke the functionality of the GPU(s). Suitable GPUs may be commercially available from vendors such as NVIDIA Corporation, ATI Technologies (AMD), and others.
System memory 1220 may store program instructions and/or data accessible by processor 1210. In various embodiments, system memory 1220 may be implemented using any suitable memory technology, such as static random access memory (SRAM), synchronous dynamic RAM (SDRAM), nonvolatile/Flash-type memory, or any other type of memory. In the illustrated embodiment, program instructions and data implementing desired functions, such as ratio mask post-filtering for audio enhancement as described above are shown stored within system memory 1220 as program instructions 1225 and data storage 1235, respectively. In other embodiments, program instructions and/or data may be received, sent or stored upon different types of computer-accessible media or on similar media separate from system memory 1220 or computer system 1200. Generally speaking, a non-transitory, computer-readable storage medium may include storage media or memory media such as magnetic or optical media, e.g., disk or CD/DVD-ROM coupled to computer system 1200 via I/O interface 1230. Program instructions and data stored via a computer-readable medium may be transmitted by transmission media or signals such as electrical, electromagnetic, or digital signals, which may be conveyed via a communication medium such as a network and/or a wireless link, such as may be implemented via network interface 1240.
In one embodiment, I/O interface 1230 may coordinate I/O traffic between processor 1210, system memory 1220, and any peripheral devices in the device, including network interface 1240 or other peripheral interfaces, such as input/output devices 1250. In some embodiments, I/O interface 1230 may perform any necessary protocol, timing or other data transformations to convert data signals from one component (e.g., system memory 1220) into a format suitable for use by another component (e.g., processor 1210). In some embodiments, I/O interface 1230 may include support for devices attached through various types of peripheral buses, such as a variant of the Peripheral Component Interconnect (PCI) bus standard or the Universal Serial Bus (USB) standard, for example. In some embodiments, the function of I/O interface 1230 may be split into two or more separate components, such as a north bridge and a south bridge, for example. In addition, in some embodiments some or all of the functionality of I/O interface 1230, such as an interface to system memory 1220, may be incorporated directly into processor 1210.
Network interface 1240 may allow data to be exchanged between computer system 1200 and other devices attached to a network, such as other computer systems, or between nodes of computer system 1200. In various embodiments, network interface 1240 may support communication via wired or wireless general data networks, such as any suitable type of Ethernet network, for example; via telecommunications/telephony networks such as analog voice networks or digital fiber communications networks; via storage area networks such as Fibre Channel SANs, or via any other suitable type of network and/or protocol.
Input/output devices 1250 may, in some embodiments, include one or more display terminals, keyboards, keypads, touchpads, scanning devices, voice or optical recognition devices, or any other devices suitable for entering or retrieving data by one or more computer system 1200. Multiple input/output devices 1250 may be present in computer system 1200 or may be distributed on various nodes of computer system 1200. In some embodiments, similar input/output devices may be separate from computer system 1200 and may interact with one or more nodes of computer system 1200 through a wired or wireless connection, such as over network interface 1240.
As shown in
Those skilled in the art will appreciate that computer system 1200 is merely illustrative and is not intended to limit the scope of the techniques as described herein. In particular, the computer system and devices may include any combination of hardware or software that can perform the indicated functions, including a computer, personal computer system, desktop computer, laptop, notebook, or netbook computer, mainframe computer system, handheld computer, workstation, network computer, a camera, a set top box, a mobile device, network device, internet appliance, PDA, wireless phones, pagers, a consumer device, video game console, handheld video game device, application server, storage device, a peripheral device such as a switch, modem, router, or in general any type of computing or electronic device. Computer system 1200 may also be connected to other devices that are not illustrated, or instead may operate as a stand-alone system. In addition, the functionality provided by the illustrated components may in some embodiments be combined in fewer components or distributed in additional components. Similarly, in some embodiments, the functionality of some of the illustrated components may not be provided and/or other additional functionality may be available.
Those skilled in the art will also appreciate that, while various items are illustrated as being stored in memory or on storage while being used, these items or portions of them may be transferred between memory and other storage devices for purposes of memory management and data integrity. Alternatively, in other embodiments some or all of the software components may execute in memory on another device and communicate with the illustrated computer system via inter-computer communication. Some or all of the system components or data structures may also be stored (e.g., as instructions or structured data) on a computer-accessible medium or a portable article to be read by an appropriate drive, various examples of which are described above. In some embodiments, instructions stored on a non-transitory, computer-accessible medium separate from computer system 1200 may be transmitted to computer system 1200 via transmission media or signals such as electrical, electromagnetic, or digital signals, conveyed via a communication medium such as a network and/or a wireless link. Various embodiments may further include receiving, sending or storing instructions and/or data implemented in accordance with the foregoing description upon a computer-accessible medium. Accordingly, the present invention may be practiced with other computer system configurations.
It is noted that any of the distributed system embodiments described herein, or any of their components, may be implemented as one or more web services. In some embodiments, a network-based service may be implemented by a software and/or hardware system designed to support interoperable machine-to-machine interaction over a network. A network-based service may have an interface described in a machine-processable format, such as the Web Services Description Language (WSDL). Other systems may interact with the web service in a manner prescribed by the description of the network-based service's interface. For example, the network-based service may describe various operations that other systems may invoke, and may describe a particular application programming interface (API) to which other systems may be expected to conform when requesting the various operations.
In various embodiments, a network-based service may be requested or invoked through the use of a message that includes parameters and/or data associated with the network-based services request. Such a message may be formatted according to a particular markup language such as Extensible Markup Language (XML), and/or may be encapsulated using a protocol such as Simple Object Access Protocol (SOAP). To perform a web services request, a network-based services client may assemble a message including the request and convey the message to an addressable endpoint (e.g., a Uniform Resource Locator (URL)) corresponding to the web service, using an Internet-based application layer transfer protocol such as Hypertext Transfer Protocol (HTTP).
In some embodiments, web services may be implemented using Representational State Transfer (“RESTful”) techniques rather than message-based techniques. For example, a web service implemented according to a RESTful technique may be invoked through parameters included within an HTTP method such as PUT, GET, or DELETE, rather than encapsulated within a SOAP message.
The various methods as illustrated in the FIGS. and described herein represent example embodiments of methods. The methods may be implemented in software, hardware, or a combination thereof. The order of method may be changed, and various elements may be added, reordered, combined, omitted, modified, etc.
Various modifications and changes may be made as would be obvious to a person skilled in the art having the benefit of this disclosure. It is intended that the invention embrace all such modifications and changes and, accordingly, the above description to be regarded in an illustrative rather than a restrictive sense.
Number | Name | Date | Kind |
---|---|---|---|
6002775 | Wood et al. | Dec 1999 | A |
8073144 | Henn et al. | Dec 2011 | B2 |
20080031462 | Walsh | Feb 2008 | A1 |
20090304203 | Haykin | Dec 2009 | A1 |
20190387346 | De Burgh | Dec 2019 | A1 |
20220406326 | Port | Dec 2022 | A1 |
Entry |
---|
International Search Report and Written Opinion mailed Sep. 4, 2023 in PCT/US2023/069049, Amazon Technologies, Inc., pp. 1-10. |
Bahareh Tolooshams et al., “A Training Framework for Stereo-Aware Speech Enhancement Using Deep Neural Networks”, 2022 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), May 23, 2022, pp. 6962-6966, IEEE. |
Number | Date | Country | |
---|---|---|---|
20240007817 A1 | Jan 2024 | US |