Abstract: Implementations relate to techniques for providing context-dependent search results. A computer-implemented method includes receiving an audio stream at a computing device during a time interval, the audio stream comprising user speech data and background audio, separating the audio stream into a first substream that includes the user speech data and a second substream that includes the background audio, identifying concepts related to the background audio, generating a set of terms related to the identified concepts, influencing a speech recognizer based on at least one of the terms related to the background audio, and obtaining a recognized version of the user speech data using the speech recognizer.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for processing queries made up of images. In one aspect, a method includes indexing images by image descriptors. The method further includes associating descriptive n-grams with the images. In another aspect, a method includes receiving a query, identifying text describing the query, and performing a search according to the text identified for the query.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for intelligently cancelling user inputs. In one aspect, a requests input by a user is received by a dialog engine. A prompt or notification regarding the request is output by the dialog engine. That the user has taken an action in response to the prompt or notification is determined by the dialog engine. Based on the action taken by the user, that the response corresponds to a potential cancellation command is determined by the dialog system.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for training a dialog system using user feedback. According to one implementation, a method includes receiving, by a dialog engine, a first input that specifies a question; providing, by the dialog engine, an answer to the question; receiving, by the dialog engine, a second input; and determining, by the dialog engine, that the second input is classified as feedback to the answer, then determining a feedback score associated with the second input.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for intelligently cancelling user inputs. In one aspect, a requests input by a user is received by a dialog engine. A prompt or notification regarding the request is output by the dialog engine. That the user has taken an action in response to the prompt or notification is determined by the dialog engine. Based on the action taken by the user, that the response corresponds to a potential cancellation command is determined by the dialog system.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for resolving ambiguity in received voice queries. An original voice query is received following one or more earlier voice queries, wherein the original voice query includes a pronoun or phrase. In one implementation, a plurality of acoustic parameters is identified for one or more words in the original voice query. A concept represented by the pronoun is identified based on the plurality of acoustic parameters, wherein the concept is associated with a particular query of the one or more earlier queries. The concept is associated with the pronoun. Alternatively, a concept may be associated with a phrase by using grammatical analysis of the query to relate the phrase to a concept derived from a prior query.

Abstract: A device may determine whether a user is facing a display screen associated with the device; and present feedback to the user. When presenting the feedback, the device may present visual information, that is based on the feedback, when determining that the user is facing the display screen associated with the device, and present audio information, that is based on the feedback, when determining that the user is not facing the display screen associated with the device. At least a portion of the audio information might not be presented when the visual information is presented when the user is facing the display screen associated with the device.

Abstract: A system is configured to obtain a first voice request, from a client, to access a voice processing system that processes voice communications received from clients; determine a first pace at which terms, associated with the first voice request, are spoken by a user of the client; determine a second pace, associated with the user, based on terms, associated with other voice requests, spoken by the user and users of the clients prior to receiving the first voice request by using a weighted average of the pace associated with the user of the client and a pace associated with the users of the clients other than the client; compare the first pace to the second pace; determine a third pace based on the comparison; and send, to the client, a voice response to be outputted at the third pace.

Abstract: Methods, systems, and apparatus, including computer program products, for providing query results using biometric parameters. One of the methods includes providing a search result in response to receiving a search query. If one or more of biometric parameters of a user indicate likely negative engagement by the user with the first search result, an additional search result is obtained and provided in response to the search query.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for processing queries made up of images. In one aspect, a method includes indexing images by image descriptors. The method further includes associating descriptive n-grams with the images. In another aspect, a method includes receiving a query, identifying text describing the query, and performing a search according to the text identified for the query.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for ranking object search results. In one aspect, a method includes providing an image to an image processing apparatus, the image defined within a first bounded region. Object search results are received from the image processing apparatus. Each object search result identifies an object belonging to one of a plurality of pre-specified object classes, and defining a respective second bounded region in which the identified object is located, the respective second bounded region being within the first bounded region. For each object search result, a respective rank score from characteristics of the respective second bounded region defined by the object search result is determined. The object search results are ranked according to their respective rank scores.

Abstract: Implementations relate to techniques for providing context-dependent search results. A computer-implemented method includes receiving an audio stream at a computing device during a time interval, the audio stream comprising user speech data and background audio, separating the audio stream into a first substream that includes the user speech data and a second substream that includes the background audio, identifying concepts related to the background audio, generating a set of terms related to the identified concepts, influencing a speech recognizer based on at least one of the terms related to the background audio, and obtaining a recognized version of the user speech data using the speech recognizer.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for processing queries made up of images. In one aspect, a method includes indexing images by image descriptors. The method further includes associating descriptive n-grams with the images. In another aspect, a method includes receiving a query, identifying text describing the query, and performing a search according to the text identified for the query.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for ranking object search results. In one aspect, a method includes providing an image to an image processing apparatus, the image defined within a first bounded region. Object search results are received from the image processing apparatus. Each object search result identifies an object belonging to one of a plurality of pre-specified object classes, and defining a respective second bounded region in which the identified object is located, the respective second bounded region being within the first bounded region. For each object search result, a respective rank score from characteristics of the respective second bounded region defined by the object search result is determined. The object search results are ranked according to their respective rank scores.

Abstract: Disclosed are systems, methods, and devices for interfacing with a wearable heads-up display via a touch-operable input device. The wearable heads-up display may include a display element for receiving and displaying display information received from a processor, and may also include a wearable frame structure supporting the display element and having a side-arm extending away from the display element. In some embodiments, the display information may appear at least partially curved to a user. In some embodiments, only a portion of the display information is shown on the at least one display element. The side-arm may be configured to secure the heads-up display to a user's body in a manner such that the display element is disposed within a field of view of the user.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for performing speech recognition using models that are based on where, within a building, a speaker makes an utterance are disclosed. The methods, systems, and apparatus include actions of receiving data corresponding to an utterance, and obtaining location indicia for an area within a building where the utterance was spoken. Further actions include selecting one or more models for speech recognition based on the location indicia, wherein each of the selected one or more models is associated with a weight based on the location indicia. Additionally, the actions include generating a composite model using the selected one or more models and the respective weights of the selected one or more models. And the actions also include generating a transcription of the utterance using the composite model.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for processing queries made up of images. In one aspect, a method includes indexing images by image descriptors. The method further includes associating descriptive n-grams with the images. In another aspect, a method includes receiving a query, identifying text describing the query, and performing a search according to the text identified for the query.

Abstract: Systems and methods for selecting an action associated with a power state transition of a head-mounted display (HMD) in the form of eyeglasses are disclosed. A signal may be received from a sensor on a nose bridge of the eyeglasses indicating if the HMD is in use. Based on the received signal, a first power state for the HMD may be determined. Responsive to the determined first power state, an action associated with a power state transition of the HMD from an existing power state to the first power state may be selected. The action may be selected from among a plurality of actions associated with a plurality of state transitions. Also, the action may be a sequence of functions performed by the HMD including modifying an operating state of a primary processing component of the HMD and a detector of the HMD configured to image an environment.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for performing speech recognition using models that are based on where, within a building, a speaker makes an utterance are disclosed. The methods, systems, and apparatus include actions of receiving data corresponding to an utterance, and obtaining location indicia for an area within a building where the utterance was spoken. Further actions include selecting one or more models for speech recognition based on the location indicia, wherein each of the selected one or more models is associated with a weight based on the location indicia. Additionally, the actions include generating a composite model using the selected one or more models and the respective weights of the selected one or more models. And the actions also include generating a transcription of the utterance using the composite model.

Abstract: A method and apparatus for enabling user query disambiguation based on a user context of a mobile computing device. According to embodiments of the invention, a first user search query, along with sensor data, is received from a mobile computing device. A recognition process is performed on the sensor data to identify at least one item. In response to determining the at least one item is a result for the first search query, data identifying the at least one item is transmitted to the mobile computing device as a response to the first search query. In response to determining the at least one item is not the result for the first search query, search results of a second search query is transmitted to the mobile computing device as the response to the first search query, the second search query comprising a query of the at least one item.