In this scenario, Contoso has previous event recordings (audio and video) in. If the content is in an audio file, it needs to be in a supported format. Use Speech Studio, Azure Speech SDK, Speech CLI, or the REST API to generate transcripts for spoken sentences and utterancesĪzure Speech provides SDKs, a CLI interface, and a REST API for generating transcripts from audio files or directly from microphone input. Operationalize the model building, evaluation, and deployment process.ġ.Use the custom model for transcription.Train, test and evaluate, and deploy the model.Ensure the data is in an acceptable format.Decide whether one or many custom models will work better. Review various options for creating custom models.If certain domain-specific words are transcribed incorrectly, consider creating a custom speech model for that specific domain.Compare the generated transcript with the human-generated transcript.Use Speech Studio, Azure Speech SDK, Speech CLI, or the REST API to generate transcripts for spoken sentences and utterances.To develop a custom speech-based application, you generally need to complete these steps: Speech service supports various languages and two fluency modes: conversational and dictation. You can create one on the Azure portal.ĭevelop a custom speech-based applicationĪ speech-based application uses the Azure Speech SDK to connect to the Azure Speech service to generate text-based audio transcription. The transcripts represent commentaries from different sports and diverse commentators. Human-generated transcripts for previous Olympics events.This article describes the application development process for this scenario: providing subtitles for an application that needs to deliver accurate event transcription.Ĭontoso already has these required prerequisite components in place: In addition, each individual sport has specific terminology that can make transcription difficult. Contoso employs female and male commentators from around the world who speak with diverse accents. As part of the broadcast agreement, Contoso provides event transcription for accessibility and data mining.Ĭontoso wants to use the Azure Speech service to provide live subtitling and audio transcription for Olympics events. This article is based on the following fictional scenario:Ĭontoso, Ltd., is a broadcast media company that airs broadcasts and commentary on Olympics events.
It provides another alternative for creating and training datasets and for operationalizing your processes. Speech CLI is a command-line tool for using Speech service without having to write any code.You can also use it to operationalize your model creation, evaluation, and deployment. Speech-to-text REST API is an API that you can use to upload your own data, test and train a custom model, compare accuracy between models, and deploy a model to a custom endpoint.It's also used to review training results. Here, it's one alternative for training datasets. Speech Studio is a set of UI-based tools for building and integrating features from Cognitive Services Speech service into your applications.Azure Cognitive Services is a set of APIs, SDKs, and services that can help you make your applications more intelligent, engaging, and discoverable.Azure Machine Learning is an enterprise-grade service for the end-to-end machine learning lifecycle.Use the APIs and CLI to help operationalize the model building, evaluation, and deployment process.Otherwise, use Speech Studio to review the word error rate (WER) and specific error details and determine what additional data is needed for training. If the performance of the custom model meets your quality expectations, publish it for use in speech transcription.Evaluate the newly trained model against the test dataset.In Speech Studio or via the API or CLI, use the new dataset to train a custom speech model.In the API and the CLI, you can pass the dataset's URI an input to create a dataset for model training and testing. Alternatively, if the dataset is in a blob store, you can use Azure Speech-to-text API and the Speech CLI. After the datasets are ready, upload them by using Speech Studio.You can combine multiple cleaned-up transcripts into one to create one dataset. Normalize the text by removing any punctuation, separating repeated words, and spelling out any large numerical values.If the transcripts are in WebVTT or SRT format, clean the files so that they include only the text portions of the transcripts.Collect existing transcripts to use to train a custom speech model.Go to part one of this guide Architectureĭownload a Visio file of this architecture.
0 kommentar(er)
