Abstract: In one embodiment, a method is disclosed comprising: analyzing, by a process, real-time video of a user participating on a video conference; determining, by the process and based on analyzing, a level of engagement of the user to the video conference; detecting, by the process, that the level of engagement of the user to the video conference is below a given threshold of engagement for a minimum length of time, wherein the minimum length of time is configurable based on one or more engagement indicators; and enacting, by the process, one or more video conference privacy measures for the user within the video conference in response to the level of engagement being below the given threshold of engagement for the minimum length of time.

Abstract: Systems, methods, and devices are disclosed for cognitive collaboration systems on a hybrid node. A query is received by a virtual assistant running on a public cloud, and it is determined whether the query pertains to data available on a public cloud resource, or the query pertains to data available on a private cloud resource. When it is determined that the query pertains to the data available on the public cloud resource, the query is interpreted by using a first model trained on at least one machine learning technique on data from the public cloud. When it is determined that the query pertains to the data available on the private cloud resource, the query is interpreted by using a second model trained on at least one machine learning technique on the data from the private cloud.

Abstract: Presented herein are techniques in which a first device connects to a communication session in which a plurality of devices communicates. The plurality of devices includes the first device and a second device. The first device outputs first audio that includes a first audio watermark associated with the communication session and the second device outputs second audio that includes a second audio watermark associated with the communication session.

Abstract: Techniques for performing conditional or controlled noise removal from audio that may contain background noise. The techniques involve obtaining audio from an environment that may have one or more unwanted noise sources, and converting the audio to digital audio data. The digital audio data is analyzed to detect whether there is noise in the audio. When noise is detected in the audio, noise removal is performed on the digital audio data to remove the noise from the audio. When noise is not detected in the audio, the digital audio data is further processed without performing noise removal on the digital audio data.

Abstract: Systems, methods, and devices are disclosed for training a model. Media data is separated into one or more clusters, each cluster based on a feature from a first model. The media data of each cluster is sampled and, based on an analysis of the sampled media data, an accuracy of the media data of each cluster is determined. The accuracy is associated with the feature from the first model. Based on a subset dataset of the media data being outside a threshold accuracy, the subset dataset is automatically forwarded to a crowd source service. Verification of the subset dataset is received from the crowd source service, and the verified subset dataset is added to the first model.

Abstract: In one embodiment, an illustrative method herein may comprise: receiving, at a receiver, an audio codec stream; determining, by the receiver, a length of time associated with a look-ahead buffer of the audio codec stream; inputting, by the receiver, the audio codec stream into an audio enhancement model trained with one or more audio enhancements to cause the audio enhancement model to apply the one or more audio enhancements to the audio codec stream to generate an enhanced audio codec stream within the length of time associated with the look-ahead buffer of the audio codec stream; and outputting, by the receiver, the enhanced audio codec stream.

Abstract: Systems, methods, and devices are disclosed for training a model. Media data is separated into one or more clusters, each cluster based on a feature from a first model. The media data of each cluster is sampled and, based on an analysis of the sampled media data, an accuracy of the media data of each cluster is determined. The accuracy is associated with the feature from the first model. Based on a subset dataset of the media data being outside a threshold accuracy, the subset dataset is automatically forwarded to a crowd source service. Verification of the subset dataset is received from the crowd source service, and the verified subset dataset is added to the first model.

Abstract: Systems, methods, and devices are disclosed for cognitive collaboration systems on a hybrid node. A query is received by a virtual assistant running on a public cloud, and it is determined whether the query pertains to data available on a public cloud resource, or the query pertains to data available on a private cloud resource. When it is determined that the query pertains to the data available on the public cloud resource, the query is interpreted by using a first model trained on at least one machine learning technique on data from the public cloud. When it is determined that the query pertains to the data available on the private cloud resource, the query is interpreted by using a second model trained on at least one machine learning technique on the data from the private cloud.

Abstract: A system and method for training a virtual assistant to recognize and learn new context for known terms is presented. The method includes receiving a natural language input, corresponding to at least one of a desired intent and a desired entity, at a natural language processor. The method involves scoring known intents based on the natural language input to generate an intent confidence score for each known intent, and scoring known entities based on the natural language input to generate an entity confidence score for each known entity. The method involves comparing the intent confidence scores and entity confidence scores to a threshold value, and determining that the natural language input does not correspond to at least one of the known intents and the known entities based on the comparing. Finally, at least one of a new intent and a new entity are determined based on the natural language input.

Abstract: Techniques for performing conditional or controlled noise removal from audio that may contain background noise. The techniques involve obtaining audio from an environment that may have one or more unwanted noise sources, and converting the audio to digital audio data. The digital audio data is analyzed to detect whether there is noise in the audio. When noise is detected in the audio, noise removal is performed on the digital audio data to remove the noise from the audio. When noise is not detected in the audio, the digital audio data is further processed without performing noise removal on the digital audio data.

Abstract: The present technology can receive audio segments from sources within one or more conference room, and can create audio fingerprints from the sources. The audio fingerprints are optimized for audio in conference room environments, which include distortions from room impulse responses, and various encoding used by telecommunication networks. In some embodiments, when two audio segments are matched, a user equipment can be instructed to mute its speakers to avoid feedback. In some embodiments, when two audio segments are matched, a user equipment can be given instructions to join a conference taking place in the room in when the audio segment originated.

Abstract: Presented herein are techniques and systems associated with generating a predicted utilization likelihood for a shared collaboration resource. Integrated resource data associated with a meeting scheduled for a shared collaboration resource is obtained and analyzed using a machine-learned predictive model. The analysis generates a predicted utilization likelihood of the shared resource. An indication of the predicted utilization likelihood is provided to an output system, such as a graphical user interface.

Abstract: The present technology pertains to a non-transitory computer readable medium comprising instructions stored thereon, the instructions effective to cause at least one processor to detect a facial recognition trigger by a device in a meeting room and present a notification in a meeting room that a camera is be to activated in order to recognize a user, wherein the meeting room is a multi-user meeting room.

Abstract: A method for communicating in a digital conversation is implemented on a computing device and includes: receiving an interactive contextual emoji from a first digital conversation participant to post in the digital conversation with at least a second digital conversation participant, where the interactive contextual emoji is pre-defined to indicate at least a current availability status associated with the first digital conversation participant, requesting the current availability status from a status application based on at least an indication of the interactive contextual emoji, where the status application maintains the current availability status, receiving the current availability status from the status application, and displaying the interactive contextual emoji in the digital conversation with at least an indication of the current availability status.

Abstract: The disclosed technology addresses the need in the art for a conference room device configured to aid the initiation, organization, and management of a conference meeting. A conference room device is configured to determine that a user is within a threshold distance of a conference room device and that the user is associated with a conference meeting at a current time. The conference room device is further configured to prompt the user to start the conference meeting, receive a confirmation from the user; and start the conference meeting.

Abstract: In one embodiment, a collaboration node prioritizes each modality of communication accessible by at least a first user and a second user based on one or more communication characteristics in a collaboration profile, monitors communication characteristics of a communication session conducted in a first modality of communication between the first user and the second user, and determines a second modality of communication accessible to the first user and the second user having a higher priority than the first modality of communication based on the collaboration profile and the communication characteristics for the communication session. The collaboration node further notifies at least one of the first user or the second user when the second modality of communication has the higher priority than the first modality of communication.

Abstract: A system and method for training a virtual assistant to recognize and learn new context for known terms is presented. The method includes receiving a natural language input, corresponding to at least one of a desired intent and a desired entity, at a natural language processor. The method involves scoring known intents based on the natural language input to generate an intent confidence score for each known intent, and scoring known entities based on the natural language input to generate an entity confidence score for each known entity. The method involves comparing the intent confidence scores and entity confidence scores to a threshold value, and determining that the natural language input does not correspond to at least one of the known intents and the known entities based on the comparing. Finally, at least one of a new intent and a new entity are determined based on the natural language input.

Abstract: The disclosed technology relates to a process for automatically training a machine learning algorithm to recognize a custom wake word. By using different text-to-speech services, input providing a custom wake word to a text to speech service can be used in order to generate different speech samples covering different variations in how the custom wake word can be pronounced. These samples are automatically generated and are subsequently used to train the wake word detection algorithm that will be used by the computing device to recognize and detect when the custom wake word is uttered by any user nearby a computing device for the purposes of initiating a virtual assistant. In a further embodiment, “white-listed” words (e.g different words that are pronounced similar to the custom wake word) are also identified and trained in order to minimize the occurrence of erroneously initiating the virtual assistant.

Abstract: A server or other computing device manages meetings in a virtual meeting room on behalf of a virtual meeting room owner. A request is received from an attendee to join a meeting in the virtual meeting room. A determination is made, based on configurations set by the virtual meeting room owner, whether to connect the attendee to a virtual waiting room. The attendee is connected to the virtual waiting room in accordance with the configurations set by the virtual meeting room owner.

Abstract: A server or other computing device manages meetings in a virtual meeting room on behalf of a virtual meeting room owner. A request is received from an attendee to join a meeting in the virtual meeting room. A determination is made, based on configurations set by the virtual meeting room owner, whether to connect the attendee to a virtual waiting room. The attendee is connected to the virtual waiting room in accordance with the configurations set by the virtual meeting room owner.