Abstract: A method for training a streaming automatic speech recognition student model includes receiving a plurality of unlabeled student training utterances. The method also includes, for each unlabeled student training utterance, generating a transcription corresponding to the respective unlabeled student training utterance using a plurality of non-streaming automated speech recognition (ASR) teacher models. The method further includes distilling a streaming ASR student model from the plurality of non-streaming ASR teacher models by training the streaming ASR student model using the plurality of unlabeled student training utterances paired with the corresponding transcriptions generated by the plurality of non-streaming ASR teacher models.

Abstract: A method for training a streaming automatic speech recognition student model includes receiving a plurality of unlabeled student training utterances. The method also includes, for each unlabeled student training utterance, generating a transcription corresponding to the respective unlabeled student training utterance using a plurality of non-streaming automated speech recognition (ASR) teacher models. The method further includes distilling a streaming ASR student model from the plurality of non-streaming ASR teacher models by training the streaming ASR student model using the plurality of unlabeled student training utterances paired with the corresponding transcriptions generated by the plurality of non-streaming ASR teacher models.

Abstract: A method includes obtaining a set of training samples, wherein each training sample includes a corresponding sequence of speech segments corresponding to a training utterance and a corresponding sequence of ground-truth transcriptions for the sequence of speech segments, and wherein each ground-truth transcription includes a start time and an end time of a corresponding speech segment. For each training sample in the set of training samples, the method includes processing, using a speech recognition model, the corresponding sequence of speech segments to obtain one or more speech recognition hypotheses for the training utterance; and, for each speech recognition hypothesis obtained for the training utterance, identifying a respective number of word errors relative to the corresponding sequence of ground-truth transcriptions.

Abstract: A method for training a streaming automatic speech recognition student model includes receiving a plurality of unlabeled student training utterances. The method also includes, for each unlabeled student training utterance, generating a transcription corresponding to the respective unlabeled student training utterance using a plurality of non-streaming automated speech recognition (ASR) teacher models. The method further includes distilling a streaming ASR student model from the plurality of non-streaming ASR teacher models by training the streaming ASR student model using the plurality of unlabeled student training utterances paired with the corresponding transcriptions generated by the plurality of non-streaming ASR teacher models.

Abstract: An image quality scorer machine accesses a candidate image to be analyzed for visual quality. The image quality scorer machine generates a visual quality score of the candidate image by first generating a prediction of a similarity score for the candidate image. The predicted similarly score of the candidate image may be generated by a process including inputting the candidate image into a neural network that has been trained to detect a set of image features in the candidate image and then to generate a corresponding predicted similarity score based on degrees to which the image features in the set are present in the candidate image. The image quality scorer machine derives the visual quality score based on the predicted similarity score outputted by the neural network. Accordingly, the image quality score machine may provide or store the generated visual quality score of candidate image for subsequent usage.

Abstract: An image quality scorer machine accesses a candidate image to be analyzed for visual quality. The image quality scorer machine generates a visual quality score of the candidate image by first generating a prediction of a similarity score for the candidate image. The predicted similarly score of the candidate image may be generated by a process including inputting the candidate image into a neural network that has been trained to detect a set of image features in the candidate image and then to generate a corresponding predicted similarity score based on degrees to which the image features in the set are present in the candidate image. The image quality scorer machine derives the visual quality score based on the predicted similarity score outputted by the neural network. Accordingly, the image quality score machine may provide or store the generated visual quality score of candidate image for subsequent usage.