Abstract: Systems and methods for automatically organizing narrated media content based on a media description file are provided. For example, a media description file may be received with narrated media content. The media description file may include information about the narrated media content, such as entitlement levels associations between narrated media content items. The narrated media content may be automatically labeled based on the media description file. A personalized user interface may be generated based on the labeled narrated media content and user subscription tiers.

Abstract: A system and method for controlling access to an on-device machine learning model without the use of encryption is described herein. For example, a request is received from an application executing on a device of a user. The request is to download a machine learning model to the device that enables a feature of the application, and the request includes information associated with the user and/or the device. The information is used to create an obfuscation key, and a derivative model can be generated using a reference copy of the machine learning model and the obfuscation key. The derivative model and the obfuscation key are then sent to the application. When the obfuscation key is provided to the derivative model at runtime, values derived from the obfuscation key are provided as additional inputs that enable the derivative model to function properly.

Abstract: A method is performed at a server system associated with a media streaming service. The method includes providing, a first client device, playback of a first media content item; and receiving, from the first client device, a second media content item to be associated with the first media content item. The method includes, after receiving the second media content item from the first client device: providing, a second client device, a user interface that includes graphical representations of a plurality of media content items that have been uploaded by other users of the media streaming service as related to the first media content item; receiving, from the second client device, a user selection of a graphical representation of the second media content item; and in response, providing the second media content item concurrently with playback of the first media content item.

Abstract: The various implementations described herein include methods and devices for speaker diarization. In one aspect, a method includes obtaining an audio recording and generating an embedding signal from the audio recording. The method further includes factoring the embedding signal to obtain a basis matrix and an activation matrix, including obtaining a sparse optimization of the embedding signal by minimizing a norm corresponding to the factored embedding signal. The method also includes generating a speaker log for the audio recording based on the sparse optimization of the embedding signal.

Abstract: A system, method and computer product for training a neural network system. The method comprises applying an audio signal to the neural network system, the audio signal including a vocal component and a non-vocal component. The method also comprises comparing an output of the neural network system to a target signal, and adjusting at least one parameter of the neural network system to reduce a result of the comparing, for training the neural network system to estimate one of the vocal component and the non-vocal component. In one example embodiment, the system comprises a U-Net architecture. After training, the system can estimate vocal or instrumental components of an audio signal, depending on which type of component the system is trained to estimate.

Abstract: Systems, methods, and devices for human-machine interfaces for improving machine understanding and fulfillment of utterance-based requests provided via the interfaces. Multiple candidate understandings from multiple stages of a natural language processing flow are preserved for arbitration and choosing by an arbitrator that applies arbitration rules to the plurality of candidates and chooses a single candidate for initiation of a corresponding service. In an embodiment, the arbitrator uses a media content taste profile to choose a candidate understanding for initiation of a corresponding service.

Abstract: A system, method and computer product for training a neural network system. The method comprises inputting an audio signal to the system to generate plural outputs f(X, ?). The audio signal includes one or more of vocal content and/or musical instrument content, and each output f(X, ?) corresponds to a respective one of the different content types. The method also comprises comparing individual outputs f(X, ?) of the neural network system to corresponding target signals. For each compared output f(X, ?), at least one parameter of the system is adjusted to reduce a result of the comparing performed for the output f(X, ?), to train the system to estimate the different content types. In one example embodiment, the system comprises a U-Net architecture. After training, the system can estimate various different types of vocal and/or instrument components of an audio signal, depending on which type of component(s) the system is trained to estimate.

Abstract: Systems, methods and computer program products provide dynamic segment resolution and transfer dynamic-segment metadata across media playback devices by detecting a transfer signal indicating a playback session is to be transferred from a first media playback device to a second media playback device, where the playback session contains first dynamic-segment metadata corresponding to a first dynamic-segment having a first data format. Second dynamic-segment metadata corresponding to a second dynamic-segment having a second data format is retrieved and the second media playback device is controlled to play the second dynamic-segment using the second dynamic-segment metadata, where the first data format and the second data format are different.

Abstract: Systems, methods and computer products are provided for retrieving play context. Media content associated with the media content is transmitted over a network. A first media playback device broadcasts an ultrasound code to one or more second media playback devices, where the ultrasound code corresponds to a play context code. The play context code is obtained from a second media playback device and in response the second media playback device receives a play context based on the play context code.

Abstract: A full attention mechanism of a multilingual transformer model is converted into a Longformer attention mechanism to generate a Longformer multilingual transformer model. The Longformer multilingual transformer model is finetuned to perform a summarization task based on episode-description:episode-transcript pairs, thereby generating a finetuned Longformer multilingual transformer model. The Longformer multilingual transformer model also can further be finetuned to perform a summarization task based on article-summary:full-original-article pairs. A summary of a query episode transcript can be generated using the single-finetuned Longformer multilingual transformer model and/or the double-finetuned Longformer multilingual transformer model. The multilingual transformer-based model enables systems, methods and computer products to be capable of generating multilingual abstractive summaries.

Abstract: An electronic device obtains a plurality of collections of media content items, each collection of media content items being associated with text generated by one or more users of the media-providing service. Based on how frequently a first media content item co-occurs with a first descriptor in text for respective collections of media items that include the first media content item, the electronic device generates, without user input, a new collection of media content items for a first user. The new collection of media content items corresponds to the first descriptor and includes the first media content item. The electronic device presents the new collection of media content items to the first user as a recommendation.

Abstract: A skip behavior analyzer is part of a media delivery system that allows for unbiased A/B testing of a plurality of versions of a song. The media delivery system stores a plurality of versions of a song and randomly selects, for each requesting device, a version of the song to associate with that device. Each time the device requests the song, thereafter, the media delivery system will provide the same version of the song for consistency. The media delivery system then gathers song play and skip information, calculates the differences in distribution of the skip behavior, and provides the skip information to allow a music composer to better determine which version of a song is more popular and why that is so.

Abstract: An architecture is provided for displaying a list of connected devices by generating a first list of devices connected to the same local network and scanning for devices within a defined physical range of a media playback device. A second list of devices within the defined physical range is generated and compared to the first list to create a third list of devices with overlapping identifiers. The media playback device then displays the third list of devices on its interface according to proximity for the user to select an appropriate device for media playback.

Abstract: Audio translation system includes a feature extractor and a style transfer machine learning model. The feature extractor generates for each of a plurality of source voice files one or more source voice parameters encoded as a collection of source feature vectors, and generates for each of a plurality of target voice files one or more target voice parameters encoded as a collection of target feature vectors. The style transfer machine learning model trained on the collection of source feature vectors for the plurality of source voice files and the collection of target feature vectors for the plurality of target voice files to generate a style transformed feature vector.

Abstract: An electronic device provides, to a user, a user-curated playlist, the user-curated playlist including an ordered set of media items that were added by the user. While providing a first media item in the ordered set of media items, the electronic device receives a first user input selecting an option to include recommended media items in the user-curated playlist. In response to the first user input, the electronic device updates the user-curated playlist to include a first recommended media item, the first recommended media item selected without user intervention based at least in part on attributes of the user-curated playlist. The first recommended media item is positioned in the user-curated playlist in between media items that were added to the ordered set of media items by the user.

Abstract: Systems and methods for estimating a long-term effect in the presence of unobserved confounding are provided. For example, a long-term effect of a change in user interface of a media playback application may be estimated. An experimental dataset and an observational dataset may be compiled using data from a plurality of computing devices. The observational dataset may include unobserved confounding. Using the experimental dataset, a short-term effect may be determined. Using the short-term effect and samples from the observational dataset, an instrumental variable may be computed that may be used in instrumental variable regression to estimate the long-term effect.

Abstract: An electronic device obtains a plurality of media items, including, for each media item in the plurality, a set of attributes of the media item. The device provides the set of attributes for each media item of the plurality of media items to a machine learning model that is trained to determine a pairwise similarity between respective media items in the plurality of media items and generates an acyclic graph of an output of the machine learning model that is trained to determine pairwise similarity distances between respective media items in the plurality of media items. The device clusters nodes of the acyclic graph, each node corresponding to a media item. Based on the clustering, the electronic device modifies metadata associated with a first media item in a first cluster and displays a representation of the first media item in a user interface according to the modified metadata.

Abstract: This disclosure is directed to systems and methods for managing a group session for consuming media content across a plurality of devices. In some configurations and by non-limiting example, the group session operates to synchronize playback and control of media content at the plurality of devices. In one aspect a method of simultaneously playing media content on a plurality of media playback devices for a group session is disclosed.

Abstract: A system for device discovery for social playback is disclosed. The system operates to connect a host media playback device to a media output device and broadcast a social playback session to guest media playback devices. Upon joining a social playback session, a guest media playback device may control the media playback at the host media playback device. Where the media output for the social playback session is provided by the media output device.

Abstract: A method includes retrieving a text from a database. The text corresponds to audio from a media content item that is provided by a media providing service, and the text includes a plurality of segments. The method also includes assigning a score for each segment in the text by applying the text to a trained computational model. The score corresponds to a predicted relevance of the respective segment to a narrative of the media content item. The method further includes identifying a non-narrative segment within the text using the assigned scores.