Abstract: Techniques are provided for training a target language model based at least in part on data associated with a reference language model. For example, language data utilized to train an English language model may be translated and provided as training data to train a German language model to recognize utterances provided in German. By utilizing the techniques herein, the efficiency of training a new language model may be improved due at least in part to replacing labor-intensive operations conventionally performed by specialized personnel with machine-generated data. Additionally, techniques discussed herein provide for reducing the time required for training a new language model by leveraging information associated with utterances of one language to train the new language model associated with a different language.

Abstract: Approaches provide for a voice communications device to control, refine, or otherwise manage the playback of media content in response to instructions, such as spoken instructions. For example, the voice communications device receives input data associated with a command, such as a request to begin media playback. Accounts corresponding to users associated with the command are identified and one or more refinements extracted from the input data are used to filter content, such as from respective content catalogs or via trained models associated with the users. Determined content is generated that includes content from each of the content catalogs or trained models associated with the users. Thereafter, the voice communications device can initiate media playback.

Abstract: Techniques are provided for training a language recognition model. For example, a language recognition model may be maintained and associated with a reference language (e.g., English). The language recognition model may be configured to accept as input an utterance in the reference language and to identify a feature to be executed in response to receiving the utterance. New language data (e.g., other utterances) provided in a different language (e.g., German) may be obtained. This new language data may be translated to English and utilized to retrain the model to recognize reference language data as well as language data translated to the reference language. Subsequent utterances (e.g., English utterances, or German utterances translated to English) may be provided to the updated model and a feature may be identified. One or more instructions may be sent to a user device to execute a set of instructions associated with the feature.

Abstract: Techniques are provided for training a language recognition model. For example, a language recognition model may be maintained and associated with a reference language (e.g., English). The language recognition model may be configured to accept as input an utterance in the reference language and to identify a feature to be executed in response to receiving the utterance. New language data (e.g., other utterances) provided in a different language (e.g., German) may be obtained. This new language data may be translated to English and utilized to retrain the model to recognize reference language data as well as language data translated to the reference language. Subsequent utterances (e.g., English utterances, or German utterances translated to English) may be provided to the updated model and a feature may be identified. One or more instructions may be sent to a user device to execute a set of instructions associated with the feature.

Abstract: Techniques are provided for training a target language model based at least in part on data associated with a reference language model. For example, language data utilized to train an English language model may be translated and provided as training data to train a German language model to recognize utterances provided in German. By utilizing the techniques herein, the efficiency of training a new language model may be improved due at least in part to replacing labor-intensive operations conventionally performed by specialized personnel with machine-generated data. Additionally, techniques discussed herein provide for reducing the time required for training a new language model by leveraging information associated with utterances of one language to train the new language model associated with a different language.

Abstract: Techniques described herein can be used to provide recommendations for multiple users. In particular, one or more users may interact with an interactive device to stream media content or utilize other services provided by a service provider. The users may provide commands to the interactive device to request content from a service provider. Contextual data associated with the request may be used to determine that an audience of the interactive device comprises more than one user. Based on this determination, content recommendations can be provided so that the recommendations are more likely to be suitable for the audience.

Abstract: The present disclosure provides computer-implemented systems and processes for augmenting user ratings of items, such as a rating of a song playing on a user device, by analyzing contextual information, such as user and/or device activity data associated with the device. The contextual information may be used to determine an associated intensity of the rating or feedback. The determined intensity levels can be used to weigh the associated rating events and improve the quality of item recommendations that are based on such ratings. Contextual information may indicate whether the user transitioned from another application to rate an item, device status information, and so on. In one embodiment, contextual information can be used to assess how intensely the user feels about a music item and to provide improved music recommendations, such as songs to provide in a playlist or radio station, based on the intensity.

Abstract: Described herein are online services provided to users based on the current context information of the users. In particular, a user may interact with an interactive device to stream media content from a remote service provider. The user may provide commands to the interactive device using a voice command. The interactive device may transmit requests based on such user command. In order for the service providers to provide customized services, the interactive device can be configured to capture additional contextual data indicative of an environment of the user such as ambient sound and provide such contextual data to the service provider. The service provider can utilize such contextual data to generate content recommendations suitable to the current context of the request or command or user.

Abstract: Approaches provide for using a voice communications device to control, refine, or otherwise manage the playback of media content in response to a spoken instruction. For example, the voice communications device can receive a request to refine and/or initiate the playback of media content, such as music, news, audio books, audio broadcasts, and other such content. Audio input data that includes the request can be received by the voice communications device and an application executing on the voice communications device or otherwise in communication with the voice communications device can analyze the audio input data to determine how to carry out the request. The application can determine whether there is an active play queue of media content configured to play using the voice communications device. In the situation where there is no media content being played using the voice communications device, the application can determine media content using information in the request.

Abstract: A technology for propagating themes is provided. In one example, a method may include identifying media having a content feature contained in a presentation of the media. The method may include extracting the content feature of the media as a continuous variable and discretizing the continuous variable into a bucket representing a discrete value. A theme label may be applied to the media and may be propagated to other media with continuous variables discretized into the subset of buckets.

Abstract: The present disclosure provides computer-implemented systems and processes for augmenting user ratings of items, such as a rating of a song playing on a user device, by analyzing contextual information, such as user and/or device activity data associated with the device. The contextual information may be used to determine an associated intensity of the rating or feedback. The determined intensity levels can be used to weigh the associated rating events and improve the quality of item recommendations that are based on such ratings. Contextual information may indicate whether the user transitioned from another application to rate an item, device status information, and so on. In one embodiment, contextual information can be used to assess how intensely the user feels about a music item and to provide improved music recommendations, such as songs to provide in a playlist or radio station, based on the intensity.

Abstract: A computing device can periodically acquire a current track time for a media content item being played. The device can periodically compare the current track time with a respective start time and a respective end time for each line (or segment) in a plurality of lines of text (or segments of text) associated with the media content item. If the device determines that the current track time is between start and end times for a given line of text, then the device can display the given line of text. A user of the device can scroll through the lines of text associated with the media content item. If the user selects a particular line of text, the device can identify a start time for the selected line of text. The device can then play the media content item at the identified start time of the selected line of text.

Abstract: A technology for synchronizing text with audio includes analyzing the audio to identify voice segments in the audio where a human voice is present and to identify non-voice segments in proximity to the voice segments. Segmented text associated with the audio, having text segments, may be identified and synchronized to the voice segments.

Abstract: A media player user interface on behalf of a user associated with a user account is generated that has a social network module that includes music updates associated with other users to which the user is linked. Media player actions associated with the user occurring in the media player user interface are also captured, updates are generated and provided to social network modules in a media player user interface associated with other linked users.

Abstract: A technology for synchronizing text with audio includes analyzing the audio to identify voice segments in the audio where a human voice is present and to identify non-voice segments in proximity to the voice segments. Segmented text associated with the audio, having text segments, may be identified and synchronized to the voice segments.

Abstract: Techniques are described for automatically selecting musical content for playback based on an initial “seed” of music selected by a user in a way that seamlessly extends the user's listening experience. The initially selected seed music might be, for example, an album or a playlist. Music that follows the seed music is algorithmically selected to match the music selected by the user.

Abstract: A computing device can periodically acquire a current track time for a media content item being played. The device can periodically compare the current track time with a respective start time and a respective end time for each line (or segment) in a plurality of lines of text (or segments of text) associated with the media content item. If the device determines that the current track time is between start and end times for a given line of text, then the device can display the given line of text. A user of the device can scroll through the lines of text associated with the media content item. If the user selects a particular line of text, the device can identify a start time for the selected line of text. The device can then play the media content item at the identified start time of the selected line of text.

Abstract: A computing device can periodically acquire a current track time for a media content item being played. The device can periodically compare the current track time with a respective start time and a respective end time for each line (or segment) in a plurality of lines of text (or segments of text) associated with the media content item. If the device determines that the current track time is between start and end times for a given line of text, then the device can display the given line of text. A user of the device can scroll through the lines of text associated with the media content item. If the user selects a particular line of text, the device can identify a start time for the selected line of text. The device can then play the media content item at the identified start time of the selected line of text.

Abstract: Disclosed are various embodiments for presenting content items that are related to media items that are currently being consumed by a user. A media item in a stream of media items being presented to a client is identified. A content item that is relevant to the media item is determined. An entry is added to the entries of a relevant content listing. The entry includes a link to the content item. The entries in the relevant content listing are arranged in a time-based sequence such that the at least one entry correlates to a sequence of playback in the client of a plurality of media files. The relevant content listing is updated in a user interface that is being presented to the user.

Abstract: Disclosed are various embodiments for presenting content items that are related to media items that are currently being consumed by a user. A media item in a stream of media items being presented to a client is identified. A content item that is relevant to the media item is determined. An entry is added to the entries of a relevant content listing. The entry includes a link to the content item. The entries in the relevant content listing are arranged in a time-based sequence such that the at least one entry correlates to a sequence of playback in the client of a plurality of media files. The relevant content listing is updated in a user interface that is being presented to the user.