Abstract: A method for controlling at least one speech-controlled graphical object. The method includes rendering, on a user interface, the at least one speech-controlled graphical object, each speech-controlled graphical object having at least one parameter associated therewith; selecting a speech-controlled graphical object from amongst the at least one speech-controlled graphical object; configuring, based on the selected speech-controlled graphical object, a speech recognition engine, associated with the user interface, to identify the at least one parameter associated with the selected speech-controlled graphical object, receive a speech signal to control the identified at least one parameter, and generate, based on the speech signal, a corresponding text representation of a speech command; and controlling, based on the speech command, the selected speech-controlled graphical object. Disclosed also is a system for controlling at least one speech-controlled graphical object.

Abstract: A method for performing domain adaptation of an end-to-end (E2E) automatic speech recognition (ASR) model. The method includes: obtaining an un-adapted version of the E2E ASR model trained using a first set of transcriptions, the un-adapted version of E2E ASR model including an encoder network, a first prediction network and a joint network; using the first set of transcriptions, while keeping parameters of first prediction network fixed, to train a language model output component to match the first prediction network; using a second set of transcriptions, while keeping parameters of language model output component fixed, to fine-tune the first prediction network for obtaining a second prediction network; and generating an adapted version of the E2E ASR model, wherein the adapted version of the E2E ASR model comprises the encoder network, the second prediction network, the language model output component, and the joint network.

Abstract: A method for performing domain adaptation of an end-to-end (E2E) automatic speech recognition (ASR) model. The method includes: obtaining an un-adapted version of the E2E ASR model trained using a first set of transcriptions, the un-adapted version of E2E ASR model including an encoder network, a first prediction network and a joint network; using the first set of transcriptions, while keeping parameters of first prediction network fixed, to train a language model output component to match the first prediction network; using a second set of transcriptions, while keeping parameters of language model output component fixed, to fine-tune the first prediction network for obtaining a second prediction network; and generating an adapted version of the E2E ASR model, wherein the adapted version of the E2E ASR model comprises the encoder network, the second prediction network, the language model output component, and the joint network.

Abstract: A method for voice recognition to perform an action. The method includes receiving a voice command, identifying a first action intended word from the voice command, assigning a confidence score to the first action intended word, altering the confidence score of the first action intended word in a temporal manner, based on confidence scores of second action intended words following the first action intended word in the voice command, identifying the action when the confidence scores of the first and second action intended words reach a pre-determined confidence score associated therewith, and performing the identified action. Disclosed also is a system for voice recognition to perform an action. The system includes one or more voice-controlled devices, and a voice managing server communicably coupled to the one or more voice-controlled devices. The voice managing server for voice recognition to perform an action using the aforementioned method.

Abstract: A method for modifying user-interface using voice commands. Method includes receiving first and second voice commands of first type respectively at first and second moments of time for creating first and second user-interface elements including first and second information having at least one information attribute associated therewith; receiving third voice command of second type at third moment of time for initiating modification process for modifying first or second user-interface elements; and rendering modified user-interface element on user-interface.

Abstract: The present invention relates to a method performing a method for functional testing of a data collection system and at least one non-transitory computer-readable memory containing computer readable instructions for implementing this method. The method includes periodically sending by the message simulator to the data collection system running in production environment, batches of sets of messages, where the messages mimic normal client messages. Each message contains a batch identification common to the messages within the specific batch and a unique message counter value. The batch of messages received by the data collection system is checked in the verification service to verify whether the data collection systems operates correctly.

Abstract: The present invention relates to a method performing a method for functional testing of a data collection system and at least one non-transitory computer-readable memory containing computer readable instructions for implementing this method. The method includes periodically sending by the message simulator to the data collection system running in production environment, batches of sets of messages, where the messages mimic normal client messages. Each message contains a batch identification common to the messages within the specific batch and a unique message counter value. The batch of messages received by the data collection system is checked in the verification service to verify whether the data collection systems operates correctly.