Abstract: Systems and methods that adaptively model network traffic to predict network capacity utilization and quality of experience into the future. The adaptive model of network traffic may be used to recommend capacity upgrades based on a score expressed in a QoE space.

Abstract: A speech interface device is configured to detect an interrupt event and process a voice command without detecting a wakeword. The device includes on-device interrupt architecture configured to detect when device-directed speech is present and send audio data to a remote system for speech processing. This architecture includes an interrupt detector that detects an interrupt event (e.g., device-directed speech) with low latency, enabling the device to quickly lower a volume of output audio and/or perform other actions in response to a potential voice command. In addition, the architecture includes a device directed classifier that processes an entire utterance and corresponding semantic information and detects device-directed speech with high accuracy. Using the device directed classifier, the device may reject the interrupt event and increase a volume of the output audio or may accept the interrupt event, causing the output audio to end and performing speech processing on the audio data.

Abstract: A speech interface device is configured to detect an interrupt event and process a voice command without detecting a wakeword. The device includes on-device interrupt architecture configured to detect when device-directed speech is present and send audio data to a remote system for speech processing. This architecture includes an interrupt detector that detects an interrupt event (e.g., device-directed speech) with low latency, enabling the device to quickly lower a volume of output audio and/or perform other actions in response to a potential voice command. In addition, the architecture includes a device directed classifier that processes an entire utterance and corresponding semantic information and detects device-directed speech with high accuracy. Using the device directed classifier, the device may reject the interrupt event and increase a volume of the output audio or may accept the interrupt event, causing the output audio to end and performing speech processing on the audio data.

Abstract: A speech interface device is configured to detect an interrupt event and process a voice command without detecting a wakeword. The device includes on-device interrupt architecture configured to detect when device-directed speech is present and send audio data to a remote system for speech processing. This architecture includes an interrupt detector that detects an interrupt event (e.g., device-directed speech) with low latency, enabling the device to quickly lower a volume of output audio and/or perform other actions in response to a potential voice command. In addition, the architecture includes a device directed classifier that processes an entire utterance and corresponding semantic information and detects device-directed speech with high accuracy. Using the device directed classifier, the device may reject the interrupt event and increase a volume of the output audio or may accept the interrupt event, causing the output audio to end and performing speech processing on the audio data.

Abstract: Systems and methods that adaptively model network traffic to predict network capacity utilization and quality of experience into the future. The adaptive model of network traffic may be used to recommend capacity upgrades based on a score expressed in a QoE space.

Abstract: A system (100) includes an update server (110) in communication with one or more client devices (106,107,108) across a network (102). A smart scheduler engine (118) schedules deployment of an update (117) to at least a first group of client devices. A monitoring auto configuration server interface (111) deploys the update to the first group of client devices. An analytics engine (114) identifies one or more anomalies occurring in the first group of client devices resulting from the upgrade. When this occurs, the scheduling engine cancels future updates to at least a second group of client devices where the later group includes devices having one or more device characteristics that correlate with other devices of the first group of client devices experiencing the one or more anomalies.

Abstract: A method of generating references to related videos is provided. Closed caption text of a primary video is analyzed to identify at least one keyword contained within the closed captioned text and a separate pre-determined listing of keywords. A keyword identified within the closed caption text and a context thereof is compared to keyword-context pairings provided within the listing. Information of a reference video related to the primary video is obtained by taking actions required by a rule in the listing associated with a matched keyword-context pairing when the keyword identified from the primary video and the context thereof is determined to match one of the keyword-context pairings in the listing. An annotation of the reference video relative to the primary video is created. A video processing electronic device and at least one non-transitory computer readable storage medium having computer program instructions stored thereon for performing the method are provided.

Abstract: A method and system is provided for managing the encoding of a plurality of video streams, each of the plurality of video streams encoded by an associated one of a plurality of encoders, each of the plurality of encoders being a member of one of a plurality of encoder pools. In one embodiment, the apparatus comprises a parser, for capturing data of a plurality of encoded video streams from the plurality of encoders, an encoding analysis module, for determining, using a user-selectable temporal portion of the captured data, a video quality stress factor for each of the plurality of encoder pools, and a dashboard generator, for generating a user interface for managing the encoding of the plurality of data streams at least in part according to the determined statistical multiplexing weighting factor video quality stress factor.

Abstract: A system (100) includes an update server (110) in communication with one or more client devices (106,107,108) across a network (102). A smart scheduler engine (118) schedules deployment of an update (117) to at least a first group of client devices. A monitoring auto configuration server interface (111) deploys the update to the first group of client devices. An analytics engine (114) identifies one or more anomalies occurring in the first group of client devices resulting from the upgrade. When this occurs, the scheduling engine cancels future updates to at least a second group of client devices where the later group includes devices having one or more device characteristics that correlate with other devices of the first group of client devices experiencing the one or more anomalies.

Abstract: A method and system is provided for managing the encoding of a plurality of video streams, each of the plurality of video streams encoded by an associated one of a plurality of encoders, each of the plurality of encoders being a member of one of a plurality of encoder pools. In one embodiment, the apparatus comprises a parser, for capturing data of a plurality of encoded video streams from the plurality of encoders, an encoding analysis module, for determining, using a user-selectable temporal portion of the captured data, a video quality stress factor for each of the plurality of encoder pools, and a dashboard generator, for generating a user interface for managing the encoding of the plurality of data streams at least in part according to the determined statistical multiplexing weighting factor video quality stress factor.

Abstract: A method of generating references to related videos is provided. Closed caption text of a primary video is analyzed to identify at least one keyword contained within the closed captioned text and a separate pre-determined listing of keywords. A keyword identified within the closed caption text and a context thereof is compared to keyword-context pairings provided within the listing. Information of a reference video related to the primary video is obtained by taking actions required by a rule in the listing associated with a matched keyword-context pairing when the keyword identified from the primary video and the context thereof is determined to match one of the keyword-context pairings in the listing. An annotation of the reference video relative to the primary video is created. A video processing electronic device and at least one non-transitory computer readable storage medium having computer program instructions stored thereon for performing the method are provided.

Abstract: A multi-tier document management system is disclosed wherein a data repository (DR) tier includes a master data repository for storing a centralized body of data. A data replication store (DRS) tier includes one or more data units for storing subsets of the data from the master repository that are relevant to the needs of the end users of the data units. A data management component (DMC) tier mediates between the data repository (DR) tier and data replication store (DRS) tier, allowing for configuration management and for performing synchronization of data. The data management component (DMC) tier may also include a configuration manager for mapping data sets to applicable end users. One or more external interfaces may be provided to the end users or customer for interfacing with the various tiers.