Abstract: Disclosed herein are system, method and/or computer program product embodiments, and/or combinations and sub-combinations thereof for providing an item search service for a premises comprising a set of Internet of Things (IoT) cameras. An example embodiment operates by receiving, via a user interface of the item search service, first user input regarding an item of interest, wherein the first user input comprises one or more of speech input or text input, accessing a plurality of images of the premises captured by the set of IoT cameras, executing a machine learning model to identify one or more images in the plurality of images that include the item of interest based at least on the first user input, generating an item search result based on the identified one or more images, and providing the item search result via the user interface of the item search service.

Abstract: A method and system for use of obfuscation coding. An example method includes a computing system receiving an image that depicts a human face. The method then includes the computing system generating a modified image based on the received image, with the generating of the modified image involving inserting into the received image an obfuscation marker that (i) obfuscates the human face in a manner that prevents a person observing the modified image from identifying the human face and (ii) includes predefined symbology that is interpretable by a machine to identify the obfuscated human face. Further, the method includes using the predefined symbology in the modified image as a basis to take action based on an identity of the human face.

Abstract: In one aspect, an example method includes (i) estimating, using a skeletal detection model, a pose of an original actor for each of multiple frames of a video; (ii) obtaining, for each of a plurality of the estimated poses, a respective image of a replacement actor; (iii) obtaining replacement speech in the replacement actor's voice that corresponds to speech of the original actor in the video; (iv) generating, using the estimated poses, the images of the replacement actor, and the replacement speech, synthetic frames corresponding to the multiple frames of the video that depict the replacement actor in place of the original actor, with the synthetic frames including facial expressions for the replacement actor that temporally align with the replacement speech; and (iv) combining the synthetic frames and the replacement speech so as to obtain a synthetic video that replaces the original actor with the replacement actor.

Abstract: A method and system for dynamically reconfiguring control pairing in a system comprising a controlling device, a first controlled device, and a second controlled device. An example method includes, when the controlling device is set to control state of the first controlled device, a processor determining that the first controlled device has experienced an operational failure. Further, the example method includes, responsive to the determining, the processor engaging in dynamic reconfiguration of the system, the dynamic reconfiguration causing the controlling device to control state of a second device instead of or in addition to the controlling device controlling state of the first controlled device.

Abstract: In one aspect, an example method includes (i) receiving audio from a first client device engaged in a communication session with a second client device, the audio comprising one or more words spoken by a user of the first client device; (ii) using the audio and a facial feature model to estimate facial movement that corresponds to the one or more words spoken by the user; (iii) generating a synthetic video depicting an avatar of the user moving according to the estimated facial movement; and (iv) in response to generating the synthetic video, causing the second client device to present the synthetic video synchronized with the audio.

Abstract: Conducting user research with human users is time consuming and expensive. Provided with quality training data that represents human users having different personae, models can be trained to offer responses that can emulate different human users. The models may offer synthetic users that can extract information from and respond to inputs to the models. The inputs can include prompts and responses can include answers to the prompts. In some cases, data about human users can be used to build different synthetic user memories, which may be used to generate different prompt chains corresponding to different users. The prompt chain can be used to prompt a model to respond based on the contextual information in the synthetic user memory.

Abstract: There is significant manual work in analyzing product reviews. Even if there are sufficient resources, human reviews can be inconsistent and subjective. A product review analysis platform leveraging engineered prompts and a large language model can address some of these issues. The platform includes a pipeline to summarize reviews, produce sentiment scores to rating categories, detect negative sentiment, extract main categories tags, extract sub-categories tags within a main categories tag, and produce weekly summaries. A dashboard can be included to visualize the enriched reviews. In some cases, synthetic users may fill in data gaps. An action recommendation engine can be included to determine appropriate resolutions. In some cases, the feature vectors generated by the large language model in response to receiving an engineered prompt can be stored in a vector database along with appropriate resolutions, such that incoming reviews can be routed appropriately using the vector database.

Abstract: In one aspect, disclosed is an example method for use in connection with a media-presentation device presenting a media program via a display screen of the media-presentation device, the method including: (i) during presentation of a temporal portion of the media program, identifying an audience engagement region of presented content, wherein the identified audience engagement region is less than an entirety of a display area that corresponds to the display screen; (ii) using at least the identified audience engagement region as a basis to select a media segment from among multiple candidate media segments; (iii) identifying an upcoming media segment insertion point within the media program; and (iv) facilitating the media-presentation device inserting, at the identified media segment insertion point within the media program, the selected media segment.

Abstract: Camera equipment used for capturing a live event is expensive and may require professionals to operate the equipment. Such camera equipment is not accessible for capturing events and experiences by everyday users. Suppose media content is captured by lower-cost input devices such as mobile phones with a video camera. A production platform can use computer vision techniques to produce enhanced media content from media content captured by the lower-cost input devices. In one example, the production platform can aggregate and analyze the received media content to extract information and determine blind spots. The production platform can create or retrieve content based on the extracted information and insert the content in the blind spots. The production platform can deliver enhanced media content to an end user device, the enhanced media content may have the received media content and the inserted content.

Abstract: Disclosed herein are system, apparatus, article of manufacture, method and/or computer program product embodiments, and/or combinations and sub-combinations thereof, for processing, understanding, and defining media content. An example can include obtaining media content items of a segment of media content; generating, based on one or more signals in the media content items, one or more media content representations encoding information about the media content items; classifying a content of the segment of the media content based on the one or more media content representations, the content of the segment of the media content being classified into one or more categories of content; and matching the segment of the media content with a targeted media content item based on the one or more categories of content associated with the segment of the media content and at least one category of content associated with the targeted media content item.

Abstract: Disclosed herein are system, apparatus, article of manufacture, method and/or computer program product embodiments, and/or combinations and sub-combinations thereof, for identifying scene breaks in media content. An example method comprises segmenting media content into a sequence of units by detecting unit boundaries. One or more feature encoders are applied to generate in an embedding space a multimedia representation of features of each unit in the sequence across different media modalities. A sequence classifier is applied to identify whether a unit boundary is a scene boundary based on the multimedia representation of units in the embedding space in at least a subset of the sequence of units.

Abstract: Disclosed herein are system, apparatus, article of manufacture, method and/or computer program product embodiments, and/or combinations and sub-combinations thereof, for processing, understanding, and defining video content. An example can include determining a first set of contextual features associated with a first portion of a media content item; identifying at least one contextual feature from the first set of contextual features that is associated with one or more targeted media content items; and selecting, based on the at least one contextual feature, a first targeted media content item from the one or more targeted media content items, wherein the first targeted media content item includes content that is related to the first portion of the media content item, and wherein the first targeted media content item is selected for presentation after the first portion of the media content item.

Abstract: Disclosed herein are system, method and/or computer program product embodiments, and/or combinations and sub-combinations thereof, for reducing and/or eliminating false positive alarm triggers. Detector devices positioned throughout an environment scan for known events that are likely indicative of an alarm condition. A source device that may generate such a sound or image may imprint a fingerprint on to the event that can be detectable by the detector devices but is imperceptible to humans. Additionally, the detector devices may use directionality and known locations of source devices to determine if a triggering event occurred at a known location of a source device. An alarm suppression decision may occur at the detector device and/or relevant information may be sent from the various detectors to a central console for a final alarm decision.

Abstract: Disclosed herein are system, apparatus, article of manufacture, method and/or computer program product embodiments, and/or combinations and sub-combinations thereof, for reducing irrelevant alerts and/or network bandwidth usage from Internet of Things (IoT) cameras through the intelligent filtering out of video data. An example embodiment operates by receiving video data from an IoT camera. The embodiment further generates an ignore mask based on the video data, wherein the ignore mask specifies one or more regions in a video stream from the IoT camera that cannot trigger an alert. The embodiment then transmits the ignore mask to the IoT camera. The embodiment then receives the video stream from the IoT camera in response to transmitting the ignore mask to the IoT camera. The embodiment then determines that the video stream is associated with the alert. The embodiment may then transmit the alert to a user device in response to the determining.

Abstract: Disclosed herein are system, method and/or computer program product embodiments, and/or combinations and sub-combinations thereof for providing a multi-alert-level monitoring and alerting service for a premises comprising a set of Internet of Things (IoT) devices. An example embodiment operates by detecting that a first event has occurred based at least on first sensor data collected from a first subset of the set of IoT devices, modifying an alert level associated with the premises from a first level to a second level in response to at least the detecting, and, in response to the modification of the alert level, modifying a set of conditions used to detect whether a second event has occurred based at least on second sensor data collected from a second subset of the set of IoT devices and/or modifying a manner in which the second sensor data is generated by or collected from the second subset.

Abstract: A method for provisioning a group of devices with control-group data defining one or more functional interrelationships between the devices. The method includes providing a GUI defining an online application marketplace, the GUI specifying multiple applications available for download, each application including a different respective set of control-group data defining a different respective set of one or more functional interrelationships. Further, the method includes receiving user input selecting a given one of the applications, the selected application defining given control-group data defining a given set of one or more functional interrelationships. And the method includes responsively (i) downloading the selected application and (ii) provisioning the devices of the group with the given set of control-group data, the provisioning establishing the given set of one or more functional interrelationships between the devices of the group.

Abstract: A method and system for controlling media-content presentation. An example method includes a computing system detecting that a media-player device has moved from a home location of the media-player device to a visited location and that the media-player device is connected with a media-presentation device at the visited location. Further, the example method includes, responsive to the detecting, the computing system causing the media-player device to output, for presentation by the media-presentation device at the visited location, media content selected by the computing system based on the media-player device having moved from the home location to the visited location. For instance, the computing system could cause the media-player device to output, for presentation by the media-presentation device, a message welcoming a user to the visited location and/or information about the visited location such as information about tourist attractions at the visited location for instance.

Abstract: Disclosed herein are system, method and/or computer program product embodiments, and/or combinations and sub-combinations thereof for providing a monitoring and alerting service for a premises comprising a set of Internet of Things (IoT) devices. An example embodiment operates by analyzing one or more of user input provided to the monitoring and alerting service or sensor data collected from at least the set of IoT devices, defining, based at least on the analyzing, an event whose occurrence can be monitored for using at least one or more IoT devices in the set of IoT devices, monitoring for an occurrence of the event using at least the one or more devices in the set of IoT devices, detecting the occurrence of the event based on the monitoring and, in response to detecting the occurrence of the event, generating an alert.

Abstract: In one aspect, an example method includes (i) obtaining media, wherein the obtained media includes (a) audio representing speech and (b) video; (ii) using at least the audio representing speech as a basis to generate speech text; (iii) using at least the audio representing speech to determine starting and ending time points of the speech; and (iv) using at least the generated speech text and the determined starting and ending time points of the speech to (a) generate closed-captioning or subtitle data that includes closed-captioning or subtitle text based on the generated speech text and (b) associating the generated closed-captioning or subtitle data with the obtained media, such that the closed-captioning or subtitle text is time-aligned with the video based on the determined starting and ending time points of the speech.

Abstract: In one aspect, an example method includes (i) estimating, using a skeletal detection model, a pose of an original actor for each of multiple frames of a video; (ii) obtaining, for each of a plurality of the estimated poses, a respective image of a replacement actor; (iii) obtaining replacement speech in the replacement actor's voice that corresponds to speech of the original actor in the video; (iv) generating, using the estimated poses, the images of the replacement actor, and the replacement speech, synthetic frames corresponding to the multiple frames of the video that depict the replacement actor in place of the original actor, with the synthetic frames including facial expressions for the replacement actor that temporally align with the replacement speech; and (iv) combining the synthetic frames and the replacement speech so as to obtain a synthetic video that replaces the original actor with the replacement actor.