You can use the console to create projects and configure their speech recognition models and speech synthesis voices. You can also use the Self-Learning Platform to customize hotwords or language models to improve recognition accuracy for your specific business needs.
Log on to the console
Use your Alibaba Cloud primary account to log on to the Intelligent Speech Interaction console. You land on the Overview page by default. The left-side navigation pane includes menu items such as Overview, All Projects, Self-Learning Platform, Personalized Voice Customization, Automated Test, Service Management and Activation, and Monitoring and Statistics. The main area displays statistics cards for the total number of projects, hotwords, language model customizations, and personalized voice customization models, as well as usage statistics for services such as Short Sentence Recognition, Real-time Speech Recognition, and Audio File Recognition. You can click Service Activation and Purchase to manage your services.
Generate an AccessToken
On the Overview page, click Get AccessToken in the upper-right corner. This token is for testing only and can be used across all your projects. Using an expired token will return a 403 error code. Refresh the page to obtain a new token. For production use, you must programmatically obtain a token. For more information, see Obtain a token.
Manage projects
On the All Projects page, create and manage speech projects. For more information, see Manage projects. Log on to the Intelligent Speech Interaction console and click All Projects in the left-side navigation pane to open the project list. The page lists all created projects with information such as Project Name, AppKey, Creation Time, Project Type, and Description. For each project, available operations include Configure Project Features, Edit, and Delete Project. Click Create Project to create a project.
Hotword
Hotwords can include names of people, places, and business-specific terms. In the left-side navigation pane, choose Self-Learning Platform>Hotword to open the hotword management page. For detailed instructions, see Manage hotwords. Click Create Hotword to create a hotword group. The page lists existing hotwords with details such as Hotword Name/ID, Type, Associated Project, and Update Time. The available operations are Preview, Update, Download, and Delete.
Language model customization
This feature is available to all users free of charge. In the left-side navigation pane, choose Self-Learning Platform > Language Model Customization to open the management page. For detailed instructions, see Manage self-learning models.
Service usage
On the Overview page, view recent usage for each service, including duration, call volume, and concurrency. For more information, see Service usage.
Automated test
Create test tasks to objectively evaluate the recognition performance of your models. For more information, see Automated Test. In the left-side navigation pane of the Intelligent Speech Interaction console, click Automated Test to open the Automated Recognition Rate Test page. This page has two tabs: Test Task and Test Dataset. On the Test Task tab, you can filter tasks by Current Status and Associated Dataset. Click Create Task to create a test task. The task list displays the Task ID, Task Name, Current Status, and Update Time. Available operations include View Result, Download Result, and Start Task.
Service activation and purchase
Intelligent Speech Interaction includes the following services:
-
Speech Recognition: Includes Short Sentence Recognition, Real-time Speech Recognition, Audio File Recognition, Audio File Recognition (Express Edition), and Audio File Recognition (Off-peak Edition).
-
Speech Synthesis: Includes Speech Synthesis and Long Text-to-Speech Synthesis.
-
Speech Analysis: Includes Speaker Recognition, Sound Event Detection, Gender Recognition, and Language Identification.
-
On-device solutions.
For more information, see Service Activation and Purchase. To learn about billing methods and view pricing, click the corresponding buttons on the right.
Intelligent Speech Interaction services are billed daily based on Alibaba Cloud's unified billing cycle and debit rules. Charges are typically posted and deducted the following day. Ensure you have a sufficient account balance to avoid service disruptions due to overdue payments. For more information, see Billing Overview.
Log on to the Intelligent Speech Interaction console, click Service Management and Activation in the left-side navigation pane, and select the Speech Recognition tab. You can view the activation status (Commercial Edition or Trial Edition), call volume limits, and concurrency limits for each service, and take actions such as Purchase Resource Plan, Upgrade to Commercial Edition, and View Prices.
Monitoring and statistics
On the Monitoring and Statistics page, you can view the call volume and concurrency for each service. For more information, see Monitoring and Statistics. At the top of the Monitoring and Statistics page, you can switch between the Speech Recognition, Speech Synthesis, and Speech Analysis tabs. Filter the results by Service (such as Short Sentence Recognition), Project, Region, and Time Range. Click Query to retrieve the results. The main area displays two line charts: Call Statistics and QPS Statistics. These charts show call volume and QPS data for the selected service and region within the specified time frame. You can also export the chart data.