140-160 and the chat and video conference provider 110, while allowing the chat and video conference provider 110 to access the decrypted multimedia streams to perform certain processing, such as recording the meeting for the participants or generating transcripts of the meeting for the participants. End-to-end encryption may be used to keep the meeting entirely private to the participants without any worry about a chat and video conference provider 110 having access to the substance of the meeting. Any suitable encryption methodology may be employed, including key-pair encryption of the streams. For example, to provide end-to-end encryption, the meeting host's client device may obtain public keys for each of the other client devices participating in the meeting and securely exchange a set of keys to encrypt and decrypt multimedia content transmitted during the meeting. Thus, the client devices 140-160 may securely communicate with each other during the meeting. Further, in some examples, certain types of encryption may be limited by the types of devices participating in the meeting. For example, telephony devices may lack the ability to encrypt and decrypt multimedia streams. Thus, while encrypting the multimedia streams may be desirable in many instances, it is not required as it may prevent some users from participating in a meeting.

By using the example system shown in FIG. 1, users can create and participate in meetings using their respective client devices 140-180 via the chat and video conference provider 110. Further, such a system enables users to use a wide variety of different client devices 140-180 from traditional standards-based video conferencing hardware to dedicated video conferencing equipment to laptop or desktop computers to handheld devices to legacy telephony devices, etc.

Referring now to FIG. 2, FIG. 2 shows an example system 200 in which a chat and video conference provider 210 provides videoconferencing functionality to various client devices 220-250. The client devices 220-250 include two conventional computing devices 220-230, dedicated equipment for a video conference room 240, and a telephony device 250. Each client device 220-250 communicates with the chat and video conference provider 210 over a communications network, such as the internet for client devices 220-240 or the PSTN for client device 250, generally as described above with respect to FIG. 1. The chat and video conference provider 210 is also in communication with one or more authentication and authorization providers 215, which can authenticate various users to the chat and video conference provider 210 generally as described above with respect to FIG. 1.

In this example, the chat and video conference provider 210 employs multiple different servers (or groups of servers) to provide different examples of video conference functionality, thereby enabling the various client devices to create and participate in video conference meetings. The chat and video conference provider 210 uses one or more real-time media servers 212, one or more network services servers 214, one or more video room gateways 216, one or more message and presence gateways 217, and one or more telephony gateways 218. Each of these servers 212-218 is connected to one or more communications networks to enable them to collectively provide access to and participation in one or more video conference meetings to the client devices 220-250.

The real-time media servers 212 provide multiplexed multimedia streams to meeting participants, such as the client devices 220-250 shown in FIG. 2. While video and audio streams typically originate at the respective client devices, they are transmitted from the client devices 220-250 to the chat and video conference provider 210 via one or more networks where they are received by the real-time media servers 212. The real-time media servers 212 determine which protocol is optimal based on, for example, proxy settings and the presence of firewalls, etc. For example, the client device might select among UDP, TCP, TLS, or HTTPS for audio and video and UDP for content screen sharing.

The real-time media servers 212 then multiplex the various video and audio streams based on the target client device and communicate multiplexed streams to each client device. For example, the real-time media servers 212 receive audio and video streams from client devices 220-240 and only an audio stream from client device 250. The real-time media servers 212 then multiplex the streams received from devices 230-250 and provide the multiplexed stream to client device 220. The real-time media servers 212 are adaptive, for example, reacting to real-time network and client changes, in how they provide these streams. For example, the real-time media servers 212 may monitor parameters such as a client's bandwidth CPU usage, memory and network I/O) as well as network parameters such as packet loss, latency and jitter to determine how to modify the way in which streams are provided.

The client device 220 receives the stream, performs any decryption, decoding, and demultiplexing on the received streams, and then outputs the audio and video using the client device's video and audio devices. In this example, the real-time media servers do not multiplex client device 220's own video and audio feeds when transmitting streams to it. Instead, each client device 220-250 only receives multimedia streams from other client devices 220-250. For telephony devices that lack video capabilities, e.g., client device 250, the real-time media servers 212 only deliver multiplex audio streams. The client device 220 may receive multiple streams for a particular communication, allowing the client device 220 to switch between streams to provide a higher quality of service.

In addition to multiplexing multimedia streams, the real-time media servers 212 may also decrypt incoming multimedia stream in some examples. As discussed above, multimedia streams may be encrypted between the client devices 220-250 and the chat and video conference provider 210. In some such examples, the real-time media servers 212 may decrypt incoming multimedia streams, multiplex the multimedia streams appropriately for the various clients, and encrypt the multiplexed streams for transmission.

As mentioned above with respect to FIG. 1, the chat and video conference provider 210 may provide certain functionality with respect to unencrypted multimedia streams at a user's request. For example, the meeting host may be able to request that the meeting be recorded or that a transcript of the audio streams be prepared, which may then be performed by the real-time media servers 212 using the decrypted multimedia streams, or the recording or transcription functionality may be off-loaded to a dedicated server (or servers), e.g., cloud recording servers, for recording the audio and video streams. In some examples, the chat and video conference provider 210 may allow a meeting participant to notify it of inappropriate behavior or content in a meeting. Such a notification may trigger the real-time media servers to 212 record a portion of the meeting for review by the chat and video conference provider 210. Still other functionality may be implemented to take actions based on the decrypted multimedia streams at the chat and video conference provider, such as monitoring video or audio quality, adjusting or changing media encoding mechanisms, etc.

It should be appreciated that multiple real-time media servers 212 may be involved in communicating data for a single meeting and multimedia streams may be routed through multiple different real-time media servers 212. In addition, the various real-time media servers 212 may not be co-located, but instead may be located at multiple different geographic locations, which may enable high-quality communications between clients that are dispersed over wide geographic areas, such as being located in different countries or on different continents. Further, in some examples, one or more of these servers may be co-located on a client's premises, e.g., at a business or other organization. For example, different geographic regions may each have one or more real-time media servers 212 to enable client devices in the same geographic region to have a high-quality connection into the chat and video conference provider 210 via local servers 212 to send and receive multimedia streams, rather than connecting to a real-time media server located in a different country or on a different continent. The local real-time media servers 212 may then communicate with physically distant servers using high-speed network infrastructure, e.g., internet backbone network(s), that otherwise might not be directly available to client devices 220-250 themselves. Thus, routing multimedia streams may be distributed throughout the video conference system and across many different real-time media servers 212.

Turning to the network services servers 214, these servers 214 provide administrative functionality to enable client devices to create or participate in meetings, send meeting invitations, create or manage user accounts or subscriptions, and other related functionality. Further, these servers may be configured to perform different functionalities or to operate at different levels of a hierarchy, e.g., for specific regions or localities, to manage portions of the chat and video conference provider under a supervisory set of servers. When a client device 220-250 accesses the chat and video conference provider 210, it will typically communicate with one or more network services servers 214 to access their account or to participate in a meeting.

When a client device 220-250 first contacts the chat and video conference provider 210 in this example, it is routed to a network services server 214. The client device may then provide access credentials for a user, e.g., a username and password or single sign-on credentials, to gain authenticated access to the chat and video conference provider 210. This process may involve the network services servers 214 contacting an authentication and authorization provider 215 to verify the provided credentials. Once the user's credentials have been accepted, and the user has consented, the network services servers 214 may perform administrative functionality, like updating user account information, if the user has account information stored with the chat and video conference provider 210, or scheduling a new meeting, by interacting with the network services servers 214. Authentication and authorization provider 215 may be used to determine which administrative functionality a given user may access according to assigned roles, permissions, groups, etc.

In some examples, users may access the chat and video conference provider 210 anonymously. When communicating anonymously, a client device 220-250 may communicate with one or more network services servers 214 but only provide information to create or join a meeting, depending on what features the chat and video conference provider allows for anonymous users. For example, an anonymous user may access the chat and video conference provider using client device 220 and provide a meeting ID and passcode. The network services server 214 may use the meeting ID to identify an upcoming or on-going meeting and verify the passcode is correct for the meeting ID. After doing so, the network services server(s) 214 may then communicate information to the client device 220 to enable the client device 220 to join the meeting and communicate with appropriate real-time media servers 212.

In cases where a user wishes to schedule a meeting, the user (anonymous or authenticated) may select an option to schedule a new meeting and may then select various meeting options, such as the date and time for the meeting, the duration for the meeting, a type of encryption to be used, one or more users to invite, privacy controls (e.g., not allowing anonymous users, preventing screen sharing, manually authorize admission to the meeting, etc.), meeting recording options, etc. The network services servers 214 may then create and store a meeting record for the scheduled meeting. When the scheduled meeting time arrives (or within a threshold period of time in advance), the network services server(s) 214 may accept requests to join the meeting from various users.

To handle requests to join a meeting, the network services server(s) 214 may receive meeting information, such as a meeting ID and passcode, from one or more client devices 220-250. The network services server(s) 214 locate a meeting record corresponding to the provided meeting ID and then confirm whether the scheduled start time for the meeting has arrived, whether the meeting host has started the meeting, and whether the passcode matches the passcode in the meeting record. If the request is made by the host, the network services server(s) 214 activates the meeting and connects the host to a real-time media server 212 to enable the host to begin sending and receiving multimedia streams.

Once the host has started the meeting, subsequent users requesting access will be admitted to the meeting if the meeting record is located and the passcode matches the passcode supplied by the requesting client device 220-250. In some examples additional access controls may be used as well. But if the network services server(s) 214 determines to admit the requesting client device 220-250 to the meeting, the network services server 214 identifies a real-time media server 212 to handle multimedia streams to and from the requesting client device 220-250 and provides information to the client device 220-250 to connect to the identified real-time media server 212. Additional client devices 220-250 may be added to the meeting as they request access through the network services server(s) 214.

After joining a meeting, client devices will send and receive multimedia streams via the real-time media servers 212, but they may also communicate with the network services servers 214 as needed during meetings. For example, if the meeting host leaves the meeting, the network services server(s) 214 may appoint another user as the new meeting host and assign host administrative privileges to that user. Hosts may have administrative privileges to allow them to manage their meetings, such as by enabling or disabling screen sharing, muting or removing users from the meeting, assigning or moving users to the mainstage or a breakout room if present, recording meetings, etc. Such functionality may be managed by the network services server(s) 214.

For example, if a host wishes to remove a user from a meeting, they may select a user to remove and issue a command through a user interface on their client device. The command may be sent to a network services server 214, which may then disconnect the selected user from the corresponding real-time media server 212. If the host wishes to remove one or more participants from a meeting, such a command may also be handled by a network services server 214, which may terminate the authorization of the one or more participants for joining the meeting.

In addition to creating and administering on-going meetings, the network services server(s) 214 may also be responsible for closing and tearing-down meetings once they have been completed. For example, the meeting host may issue a command to end an on-going meeting, which is sent to a network services server 214. The network services server 214 may then remove any remaining participants from the meeting, communicate with one or more real time media servers 212 to stop streaming audio and video for the meeting, and deactivate, e.g., by deleting a corresponding passcode for the meeting from the meeting record, or delete the meeting record(s) corresponding to the meeting. Thus, if a user later attempts to access the meeting, the network services server(s) 214 may deny the request.

Depending on the functionality provided by the chat and video conference provider, the network services server(s) 214 may provide additional functionality, such as by providing private meeting capabilities for organizations, special types of meetings (e.g., webinars), etc. Such functionality may be provided according to various examples of video conferencing providers according to this description.

Referring now to the video room gateway servers 216, these servers 216 provide an interface between dedicated video conferencing hardware, such as may be used in dedicated video conferencing rooms. Such video conferencing hardware may include one or more cameras and microphones and a computing device designed to receive video and audio streams from each of the cameras and microphones and connect with the chat and video conference provider 210. For example, the video conferencing hardware may be provided by the chat and video conference provider to one or more of its subscribers, which may provide access credentials to the video conferencing hardware to use to connect to the chat and video conference provider 210.

The video room gateway servers 216 provide specialized authentication and communication with the dedicated video conferencing hardware that may not be available to other client devices 220-230, 250. For example, the video conferencing hardware may register with the chat and video conference provider when it is first installed and the video room gateway may authenticate the video conferencing hardware using such registration as well as information provided to the video room gateway server(s) 216 when dedicated video conferencing hardware connects to it, such as device ID information, subscriber information, hardware capabilities, hardware version information etc. Upon receiving such information and authenticating the dedicated video conferencing hardware, the video room gateway server(s) 216 may interact with the network services servers 214 and real-time media servers 212 to allow the video conferencing hardware to create or join meetings hosted by the chat and video conference provider 210.

Referring now to the telephony gateway servers 218, these servers 218 enable and facilitate telephony devices' participation in meetings hosted by the chat and video conference provider 210. Because telephony devices communicate using the PSTN and not using computer networking protocols, such as TCP/IP, the telephony gateway servers 218 act as an interface that converts between the PSTN, and the networking system used by the chat and video conference provider 210.

For example, if a user uses a telephony device to connect to a meeting, they may dial a phone number corresponding to one of the chat and video conference provider's telephony gateway servers 218. The telephony gateway server 218 will answer the call and generate audio messages requesting information from the user, such as a meeting ID and passcode. The user may enter such information using buttons on the telephony device, e.g., by sending dual-tone multi-frequency (“DTMF”) audio streams to the telephony gateway server 218. The telephony gateway server 218 determines the numbers or letters entered by the user and provides the meeting ID and passcode information to the network services servers 214, along with a request to join or start the meeting, generally as described above. Once the telephony client device 250 has been accepted into a meeting, the telephony gateway server is instead joined to the meeting on the telephony device's behalf.

After joining the meeting, the telephony gateway server 218 receives an audio stream from the telephony device and provides it to the corresponding real-time media server 212 and receives audio streams from the real-time media server 212, decodes them, and provides the decoded audio to the telephony device. Thus, the telephony gateway servers 218 operate essentially as client devices, while the telephony device operates largely as an input/output device, e.g., a microphone and speaker, for the corresponding telephony gateway server 218, thereby enabling the user of the telephony device to participate in the meeting despite not using a computing device or video.

It should be appreciated that the components of the chat and video conference provider 210 discussed above are merely examples of such devices and an example architecture. Some video conference providers may provide more or less functionality than described above and may not separate functionality into different types of servers as discussed above. Instead, any suitable servers and network architectures may be used according to different examples.

Referring now to FIGS. 3A-3B, FIG. 3A shows an example system 300 for automatic speech recognition using multiple language models. The system 300 includes a virtual conference provider 310, which can be connected to multiple client device 330, 340a-n via one or more intervening communication networks 320. In this example, the communications network 320 is the internet, however, any suitable communications network or combination of communications network may be employed, including LANs (e.g., within a corporate private LAN), WANs, etc.

Each client device 330, 340a-n executes virtual conference software that connects to the virtual conference provider 310 and joins a meeting. During the meeting, the various participants (using virtual conference software or “client software” at their respective client devices 330, 340a-n) are able to interact with each other to conduct the meeting, such as by viewing video feeds and hearing audio feeds from other participants, and by capturing and transmitting video and audio of themselves. The virtual conference provider 310 is configured to host meetings between different users and can record the audio and video streams from those meetings to provide a recording of the meeting or to generate a transcript using an example system for automatic speech recognition using multiple language models according to this disclosure.

Referring now to FIG. 3B, FIG. 3B illustrates the virtual conference provider 310 discussed above with respect to FIG. 3A. The virtual conference provider in this example has been configured with a multi-model ASR system 316 to provide automatic speech recognition using multiple language models. The multi-model ASR system 316 includes ASR functionality 350 as well as a rescoring component 360 that includes a language model 362, scoring analysis functionality 370, and a large language model 380. While each of these components is depicted as being part of the virtual conference provider 310, it should be appreciated that they may be distributed across any number of computing devices managed by the virtual conference provider 310. Further, in some examples, one or more components may be hosted by a third party that is accessible by the virtual conference provider 310. In this example, however, all components are hosted by the virtual conference provider 310.

In this example system, the ASR functionality 350 is a conventional ASR system that is configured to generate an “N-best” list, or a list of proposed transcriptions of an utterance, which are generally referred to as “hypotheses” within this disclosure. In this example, a transformer network is employed, but other suitable ASR functionalities 350 may be embodied by recurrent neural networks, long short-term memory networks, or any other suitable ASR system that can generate multiple hypotheses from an inputted utterance. An “utterance” refers to audio that includes one or more spoken words. The utterance may represent a word, phrase, sentence, thought, or other grouping of words. From the utterance, the ASR functionality 350 generates multiple hypotheses from a beam search of ASR decoding. In some examples, the ASR functionality 350 may also generate a corresponding score for each hypothesis.

The rescoring component 360 accepts the hypotheses 352 output by the ASR functionality 350 and generates a score for each using a language model 362. Any suitable language model may be used as a rescoring language model 362, such as auto-regressive language models, bidirectional encoder representation from transformers (“BERT”) models, or large language models (such as GPT-2, GPT-J, ChatGPT-3.5, or ChatGPT-4). In this example, an auto-regressive language model is employed as the rescoring language model 362.

In this example, rescoring involves generating the log-likelihood of each hypothesis within the hypotheses 352 generated by the ASR functionality 350, such as using the following equation for each hypothesis, y:

${Score}_{L M} (y) = \log P_{L M} (y) = \sum_{i = 1}^{L} P_{L M} (y_{i} | y_{< i})$

where y_<i=(y₁, . . . , y_i-1). It should be appreciated that because auto-regressive LLMs are unidirectional in the sense that the prediction of y₁only depends on the previous context, which makes it computationally efficient, only requiring a single inference pass to compute the log-likelihood in an auto-regressive manner. However, any technique for generating score indicating probabilities that each hypothesis includes a sequence of words according to a particular language, such as the softmax function.

If the ASR functionality 350 generates its own scores, the scores generated by the rescoring functionality 360 are combined with the original ASR scores for the corresponding hypotheses to generate a new score. In this example, the scores are combined using the following equation:

$Score (y) = \log P_{A S R} (y | x) + α \log P_{L M} (y)$

where x denotes acoustic features, y denotes the hypothesis from the ASR functionality 350, and a is a hyper-parameter weight. In this case, a is set to three, though any suitable value may be used. However, any suitable technique to combine the scores may be employed.

After rescoring, scoring analysis 370 identifies the hypothesis with the best score and evaluates one or more thresholds. In this example, the scoring analysis 370 compares the highest score against a threshold score. If the score is greater than the threshold score, the threshold is satisfied. The scoring analysis 370 then outputs the hypothesis as the final transcription of the inputted utterance.

In some examples, additional thresholds may be used, however. For example, another threshold may be established to evaluate a difference in score between the highest score and the second-highest score, after rescoring. The difference of these two scores may then be compared against the additional threshold, and if it is greater than the threshold, the additional threshold is satisfied. And still further thresholds may be employed, such as differences between the highest score and the average of the remaining scores or the average of the top three scores. In some examples, entropy associated with the hypotheses may be employed to evaluate the best hypothesis. Further, in some examples, the scores output by the ASR functionality 350 and the rescoring functionality 360 may be compared to separate predefined thresholds.

It should be appreciated as well that while the example above contemplates scores where a higher score is better, other scoring domains may be employed. For example, in some examples, the lowest score may be the “best” score or a score closest to zero or another value may be the best score. Thus, a threshold may be a score, a difference between two values, such as a score and a limit value, or any other suitable metric that can be compared against the output by the rescoring functionality 360. Further, while the inequality “greater than” was specified above, any suitable comparison may be employed, such as “greater than or equal,” “less than,” “less than or equal,” or whether a score is within a particular range.

If the highest score does not satisfy the threshold(s), a LLM 380 is employed. In this example, the LLM 380 comprises the ChatGPT-4 LLM, though any suitable LLM, such as ChatGPT-3.5, Language Model for Dialogue Applications (or “LaMDA”) (such as Google Bard), DeepMind Sparrow, including future versions of any of these or other LLMs, may be used. The LLM 380 receives the hypotheses generated by the ASR functionality 350 and a request to provide a corrected transcription of the best hypothesis identified before rescoring. In addition, the LLM 380 may be provided with one or more constraints 306 to assist it with generating and outputting a corrected transcription.

Because the LLM 380 employs the ChatGPT-4 LLM, which accepts natural language queries and prompts, the LLM 380 is provided the best hypothesis and the constraints 306 as natural language instructions. The constraints 306 may be predetermined and stored in a data store 314 to be later provided to the LLM 380 to assist or otherwise constrain its processing of the best hypothesis. However, it should be appreciated that “constraints” provided to the LLM 380 may not be strictly adhered to by the LLM 380 due to the particular LLM that is employed, but instead may provide guidance to the LLM 380, which may be partially or totally ignored in the course of generating a final transcription based on the best hypothesis.

Constraints 306 may include requirements to limit the words used in the final transcription to those found within the set of hypotheses output by the ASR functionality 350, maintaining the same sentence structure or number of words in the best hypothesis, ignoring punctuation, etc. For example, the following set of predetermined constraints may be provided to the LLM 380:

- I want you to check and correct potential errors in one sentence according to the following rules. Here is the sentence to work on: [best hypothesis].
- You need to first consider the following variant sentences and only pick corrected words from them: [set of hypotheses].

Here are some additional rules for this correction:

- 1. If any word in the original sentence looks weird or inconsistent, then replace it with a corresponding word from variant sentences.
- 2. You don't have to modify the original sentence if it already looks good.
- 3. Keep the sentence structure and word order intact.
- 4. Only replace words in the original sentence with ones from variant sentences. Do not simply add or delete words.
- 5. Try to make the corrected sentence have the same number of words as the original sentence.
- 6. Ignore punctuation.
- 7. Use U.S. English.
- 8. Output only one modified sentence and no need to provide explanation.
  
  However, any one or more of the constraints above may be omitted in some examples. For example, some examples may not include a constraint to ignore punctuation. In addition, while the constraints above specify U.S. English, any language or language dialect may be specified, as appropriate for a particular context.

In addition, other constraints may be provided instead or in addition to these. For example, additional constraints may include one or more specialized words 304, such as associated with a particular field, a particular project, or a particular industry or company. One or more lists of such words may be stored in data store 314 and accessed by scoring analysis 370 to provide to the LLM 380. A suitable constraint to employ such specialized words 304 may be “Only replace words in the original sentence with ones from variant sentences or from the following list of words: [specialized words]. Do not simply add or delete words.” Another constraint may be to generate a semantically identical or similar sentence without verbal pauses, or to generate a verbatim transcript including verbal pauses.

One issue with LLMs that may be addressed by using constraints 306 is that LLMs may have a propensity to “hallucinate,” such as by performing a semantic analysis of the best hypothesis and then generate words or phrases to elaborate upon the semantic content of the hypothesis, thereby generating a transcription that includes words not present in the original utterance. To help reduce the likelihood of such hallucinations, constraints such as (3) and (5) in the example constraints above may be employed, which constrain the length and vocabulary to be used in generating the transcription from the best hypothesis. Thus, the LLM 380 can be instructed to generate a transcription rather than to employ any “creativity” in generating the final transcription.

Another technique to limit the LLM's propensity to hallucinate, or to limit its “creativity,” is to adjust a “temperature” or equivalent hyperparameter of the LLM 380. Generally, temperature relates to flexibility in selecting words for use in a sequence of words, thereby allowing the use of words with lower probabilities. The lower the temperature setting for ChatGPT-4, the more likely the LLM 380 will generate outputs that are truer to a particular format or guideline, while a higher temperature setting affords more variability, flexibility, and, potentially, less coherence in its outputs. Thus, a constraint 306 supplied to the LLM 380 may include a temperature, or equivalent hyperparameter (or set of hyperparameters, depending on the LLM employed). In this example, a temperature of 0.2 (from a range of 0.0-1.0) is employed, however, any suitable temperature may be used.

After the LLM 380 has received the inputted hypotheses and constraints 306, it operates to generate and output a finalized transcription. It should be appreciated that while the multi-model ASR system 316 depicted in FIG. 3B employs a rescoring language model 362 and a LLM 380, which are described as being different models, in some examples, the LLM 380 may be used to provide rescoring as well and thus may operate as the rescoring language model 362. Thus, while the same model may be used for different purposes, the different uses of the same model can provide for improved automatic speech recognition according to this disclosure.

It should be appreciated that the architecture depicted in FIG. 3B provides an example implementation and division of functionality within a system for ASR using multiple language models. Systems according to this disclosure need not be divided or grouped as depicted in FIG. 3B. Further, different components of examples systems according to this disclosure may be provided by multiple entities. For example, a virtual conference provider