Abstract: A method of generating a trained trigger word detection model includes training an auxiliary model, based on an auxiliary task, to concentrate on one or more utterances and/or learn context of the one or more utterances using generic single word and/or phrase training data; and obtaining a trigger word detection model by retraining one or more final layers of the auxiliary model, which is weighted based on the auxiliary task, based on a trigger word detection task that detects one or more trigger words. The retraining uses training data specific to the one or more trigger words.

Abstract: A method includes obtaining an audio input and a location associated with an electronic device. The method also includes generating an audio embedding associated with the audio input. The method further includes determining a first difference between the audio embedding associated with the audio input and an audio embedding associated with a known user. The method also includes determining a second difference between the location associated with the electronic device and a known location associated with the known user. The method further includes generating, using a false trigger mitigation (FTM) system, a probability of the audio input including a false trigger for automatic speech recognition based on the audio input, the first difference, and the second difference. In addition, the method includes determining whether to perform automatic speech recognition based on the probability.

Abstract: A method comprises obtaining an audio input. The method also includes providing at least a portion of the audio input to a frame-level detector model. The method also includes obtaining a first output of the frame-level detector model including frame-level predictions associated with at least the portion of the audio input. The method also includes providing at least one chunked audio frame to a word-level verifier model. The method also includes obtaining a second output of the word-level verifier model including word-level probabilities associated with the at least one chunked audio frame. The method also includes instructing performance of automatic speech recognition on the audio input based on the word-level probabilities associated with the at least one chunked audio frame.

Abstract: A method includes obtaining audio data and identifying an utterance of a wake word or phrase in the audio data. The method also includes generating an embedding vector based on the utterance from the audio data and accessing a set of previously-generated vectors representing previous utterances of the wake word or phrase. The method further includes performing clustering on the embedding vector and the set of previously-generated vectors to identify a cluster including the embedding vector, where the identified cluster is associated with a speaker. The method also includes updating a speaker vector associated with the speaker based on the embedding vector and determining, using a speaker verification model, a similarity score between the updated speaker vector and the embedding vector. In addition, the method includes determining, based on the similarity score, whether a speaker providing the utterance matches the speaker associated with the identified cluster.

Abstract: A method includes accessing, using at least one processor of an electronic device, a machine learning model. The machine learning model is a trained student model that is trained using audio samples in a plurality of accent types. The method also includes receiving, using the at least one processor, an audio input from an audio input device. The method further includes providing, using the at least one processor, the audio input to the trained student model. The method also includes receiving, using the at least one processor, an output from the trained student model including frame-level probabilities associated with the audio input. In addition, the method includes instructing, using the at least one processor, at least one action based on the frame-level probabilities associated with the audio input.

Abstract: A method of generating a trained trigger word detection model includes training an auxiliary model, based on an auxiliary task, to concentrate on one or more utterances and/or learn context of the one or more utterances using generic single word and/or phrase training data: and obtaining a trigger word detection model by retraining one or more final layers of the auxiliary model, which is weighted based on the auxiliary task, based on a trigger word detection task that detects one or more trigger words. The retraining uses training data specific to the one or more trigger words.

Abstract: An application is operable to achieve a plurality of application states. Static analysis is performed based on machine code of the application to generate a state access instruction configured to invoke a state of the application states of the application and identify a set of parameter variables relied on by the state access instruction. Dynamic analysis is performed to determine values for the set of parameter variables to invoke the state. The state access instruction and the values determined for the set of parameter variables to invoke the state are stored as an access mechanism. The application is executed, and data is scraped from the state. The scraped data can be used to determine whether the state of the application is a relevant search result, and the access mechanism can be used to launch and transition the application to the state.

Abstract: Some aspects of the disclosure provide a method comprising obtaining machine executable code of an application, the application operable to achieve a set of application states, pre-processing the machine executable code to generate reviewable code, identifying, from the reviewable code, a set of state access instructions configured to invoke or assist in invoking one of the set of application states of the application, the set of state access instructions indicating a first state access instruction configured to invoke a first state of the set of application states and a second state access instruction configured to invoke a second state of the set of application states that is different from the first state, each of the set of state access instructions including an application resource identifier referencing an application and indicating an operation for the application to perform, and storing the set of state access instructions.

Abstract: Some aspects of the disclosure provide a method comprising obtaining machine executable code of an application, the application operable to achieve a set of application states, pre-processing the machine executable code to generate reviewable code, identifying, from the reviewable code, a set of state access instructions configured to invoke or assist in invoking one of the set of application states of the application, the set of state access instructions indicating a first state access instruction configured to invoke a first state of the set of application states and a second state access instruction configured to invoke a second state of the set of application states that is different from the first state, each of the set of state access instructions including an application resource identifier referencing an application and indicating an operation for the application to perform, and storing the set of state access instructions.