Abstract: Systems and methods for user authentication in video conferences using signed contact lists are provided. A client device associated with a user joins a video conference hosted by a video conference provider. The video conference has a plurality of participants. The client device receives participant information for each participant of the plurality of participants. The client device verifies at least one participant of the plurality of participants by comparing the participant information with a contact list associated with the user. The client device updates a contact record corresponding to the at least one participant in the contact list. The contact record includes video conference information, at least a portion of the participant information, and a cryptographic signature.

Abstract: In some aspects, techniques may include receiving media streams from one or more client devices. The media streams can be received by a virtual conference provider. Also, the techniques may include selecting a subset of the media streams based on one or more characteristics of the media streams. The streams may be selected using a machine learning (ML) model. In addition, the techniques may include identifying one or more segments of the subset of media streams satisfying an inclusion criteria. Moreover, the techniques may include generating a recording of the virtual conference including the one or more identified segments.

Abstract: Example methods and systems provide a workplace assistant application that can detect electronic content associated with a remote user and access multiple communication channels connected to a workplace assistant client application. The system can submit the electronic content to one or more predictive models to provide a context for the electronic content. The context is based at least in part on information from multiple channels. The system can generate, using the predictive model(s), an action item based at least in part on the context of the electronic content as well as display or store the action item using the workplace assistant client application. An action item, as examples, can be a “to do” item, a meeting agenda, a reminder, a meeting, a task, or some combination of one or more of these.

Abstract: Techniques for providing video conferencing metrics in autonomous systems are disclosed. In an example method, a video conference provider joins a client device to the video conference, where the client device is connected to an internet protocol (IP) network associated with a first autonomous system (AS) managed by a first communication service provider (CSP) associated with a first AS number (ASN). The video conference provider determines, from packets received from the client device, client device information. The video conference provider determines, based on the client device information, metrics about performance of the video conference in the first AS. The video conference provider outputs in real-time the metrics about the performance.

Abstract: Systems and methods for active speaker detection for videoconferencing are provided. For example, an audio recording system can record a set of raw audio signals. The set of raw audio signals include audio signals recorded by a set of microphones at different distances from an audio source with different orientations. An audio dataset generation system can access a virtual meeting room setup which specifies microphones used in a virtual meeting room, locations of speakers and the microphones, and orientations of the speakers. The audio dataset generation system generates a synthetic audio signal for each of the speakers specified in the virtual meeting room setup by combining audio signals selected from the set of raw audio signals according to the virtual meeting room setup.

Abstract: Sensor data that include ambient qualities of multiple locations within a physical environment is obtained by a device. A location within the physical environment at which to place a participant device to connect to a video conference is determined based on the sensor data. An indication of the location is output for display at the device.

Abstract: One example method includes joining, by a first client device, a videoconferencing meeting hosted by a video conference provider, the videoconference meeting including a plurality of participants; providing an audio stream and a video stream to a video conference provider; receiving, from a second client device, an audio focus area associated with a video stream provided the first client device; determining, based on the audio focus area, a bounding region within an environment shown in the video stream; directing a audio capture device to capture audio from the bounding region; and providing the captured audio as an audio stream to the video conference provider.

Abstract: Systems and methods for machine learning based audio watermarking for videoconferencing are provided. For example, a computing device accesses an original audio signal and a watermark to be embedded into the original audio signal and extracts, using an audio encoder, a set of audio features from the original audio signal. The audio encoder is a machine learning model. The computing device further extracts, using a watermark encoder, a set of watermark features from the watermark. The watermark encoder is also a machine learning model. The computing device combines the set of audio features and the set of watermark features to generate a set of features, compresses the set of features, and transmits the compressed set of features.

Abstract: Methods and systems provide for virtual meeting coaching with dynamically extracted content. In one embodiment, the system receives transcripts for a plurality of selected meetings from within a communication platform, the meetings being selected for question extraction; extracts, via a question extraction model, questions from the transcripts for the selected meetings; extracts, from the questions, a number of expected answers, each expected answer associated with one of the questions; connects to a coaching session with one or more participants and a virtual coaching agent; transmits, by the virtual coaching agent, at least a subset of the extracted questions to the participants; receives answers to the subset of the extracted questions by participants; and generates one or more evaluation scores for the participants corresponding to each of the answers, the evaluation scores being based at least in part on the extracted expected answers associated with the subset of the extracted questions.

Abstract: One example method includes receiving, by a virtual conference provider, audio streams from a plurality of participants during a virtual conference; performing speech recognition, by the virtual conference provider, on the received audio streams to generate a transcript of the virtual conference; determining one or more utterances made during the virtual conference; determining, using a first trained machine learning (“ML”) model, one or more sub-topics discussed during the virtual conference based on the one or more utterances; determining, using a second trained machine learning (“ML”) model, one or more topics discussed during the virtual conference based on the one or more sub-topics; determining metadata corresponding to each of the one or more topics, the one or more sub-topics, and the one or more utterances; and generating and providing suggested follow-up topics for a subsequent virtual conference based on the metadata.

Abstract: Systems and methods for failover monitoring using dynamic traffic tagging are disclosed. In an example method, a computing device outputs failover information including a designation of a current failover location. The computing device receives, from a server device, a failover status message indicating that the server device received a request message including the current failover location. The computing device then outputs a second failover information including a second designation of a second network location as the current failover location. The computing device receives from the server device, a second failover status message indicating that the server device received a second request message including a network location other than the current failover location. The computing device then outputs a failover failure notification and outputs a command to cause a corrective action.

Abstract: Methods and systems for transmitting content during a networked conference. In an embodiment, a method is provided for secure access to online events. The method includes receiving a request from a user to access an online event, obtaining a user email address, an event identifier and an event passcode, and hashing the user email address, the event identifier and the event passcode to generate a ticket hash value. The method also includes forming a secure ticket that comprises the ticket hash value, and transmitting the secure ticket to the user. The method also includes receiving the secure ticket from the user when the user want to access the event, hashing the user email address, the event identifier, and the event passcode to generate a confirming hash value, and comparing the two hash values and granting event access to the user if the two hash values match.

Abstract: A machine learning model (e.g., including a deep learning neural network) with learned embeddings is applied to time series data with associated metadata to obtain predictions of the time series value. For example, a call volume in a period of time may be predicted based on call volume data for a sequence of time bins in a window of preceding time. Time bins may be associated with respective metadata, such as day of week, hour of day, day of month, holiday, part of business cycle, weather, and/or tide. These pieces of metadata may be mapped to embedding vectors using trained embedding functions. The resulting embedding vectors may be input to a neural network along with the corresponding time series data (e.g., call volumes) to make a prediction for future time bin. For example, the prediction may be used to provision servers in a network infrastructure.

Abstract: A conferencing software receives a request from a device of a caller to connect to a conference. The conferencing software determines that name information of the caller is not included in the request. In response to determining that the name information is not included in the request, the conferencing software transmits an identification request to the device, where the identification request is a text-based message that is configured to obtain the name information from the caller. The conferencing software displays the name information received in a response to the identification request in a user interface associated with the conference.

Abstract: A conferencing system identifies key participants in a video conference and generates a graphical user interface (GUI) with enlarged participant tiles for displaying video streams of the identified key participants. The conferencing system outputs the GUI to client devices connected to the video conference. The conferencing system sends a notification to the client devices when a key participant joins the video conference or speaks during the video conference.

Abstract: A metadata management service receives operational data generated by operations of a computer system is received. The operational data is analyzed to identify data structures within the operational data. A schema definition that represents the operational data in a standardized format is generated based on the data structures. The schema definition is integrated into the metadata management service.

Abstract: One example method includes transmitting, by a client device, a request for a reduced large language model (“LLM”) to a remote server; receiving, by the client device from the remote server, and storing the reduced LLM, the reduced LLM based on a trained general LLM; receiving, by the client device, a request to generate content using the reduced LLM; providing the request to the reduced LLM; and receiving generated content from the reduced LLM based on the request.

Abstract: Systems and methods for automated code generation for database configuration, security, and operation are disclosed. In an example computer-implemented method, a computing system accesses database configuration information for a database. The computing system determines, based on the database configuration information, one or more database instructions to be executed by the database. The computing system generates, for each of the one or more database instructions, a program code portion for executing the database instruction. The computing system generates database client program code including the one or more program code portions. The computing system outputs the database client program code to cause execution of the database client program code.

Abstract: A liveness requirement can be enforced on a videoconference by performing some techniques described herein. For example, a system can sequentially generate strings during a videoconference. The system can sequentially transmit the strings to a host device associated with a host of the videoconference. The system can receive a communication from the host device during the videoconference. The system can determine whether the communication includes a most recently generated string among the strings. Based on determining that the communication excludes the most recently generated string, the system can discard the communication as outdated.

Abstract: Systems and methods for synthetic audio datasets generation for videoconferencing based on realistic room acoustic simulation are provided. For example, a room-independent recording for a source audio signal and a particular type of microphones can be generated and a target room setup for a target room can be obtained. The target room setup specifying one or more of a size of the target room, respective locations of a speaker and a microphone in the target room. Room characteristics of the target room can be generated based on the target room setup via a room acoustic model. A synthetic audio signal for the target room and the particular type of microphones can be generated by applying the room acoustic characteristics onto the room-independent recording.