Google has introduced the Voice Interaction API as part of the recently announced Android M. The API enables apps to interact with users via spoken dialogue. While Google has supported voice commands and text to speech for some time, the API creates a true dialogue. For example, if a user engages with a smart home app using the command "Turn on the lights," the Voice Interaction API can respond accordingly by inquiring as to which room the user means.
"I can't wait for the day when I can talk to my watch and phone like they do in the movies," Sunil Vemuri, Google's voice actions product manager, explained in an introductory YouTube video. "We're still a ways away from that. But the good news is [Google is] releasing a new API on Android M that takes us an important step in that direction."
Many anticipated the release of a Voice Access API at Google I/O 2015. However, the Voice Access API was removed from the schedule. Instead, the Voice Interaction API was released as part of Android M. Last week, Google released a full video detailing the capabilities of the Voice Interaction API and where developers can find more information.
"OK Google" constitutes the "hot word" that engages the assistant to start listening to the user. The speech to text, natural language processing and knowledge graph work together to determine what actions should ensue. If an action is needed, an Android intent is triggered. The Voice Interaction API allows apps to make a number of decisions based on user direction (e.g., mature content identification, clarifications, narrowed scope of search). System voice actions include media, search, communications and fitness. In addition to the preset system actions, Google includes custom voice actions that allow users to personalize app interaction.
During the introduction video, Vemuri demonstrates the use of the Voice Interaction API by vocally selecting music with the TuneIn app. In addition to the TuneIn Integration, Google lists a number of examples and provides Sample Code at its CodeLab. While the Voice Access API may have missed its Google I/O appearance, the Voice Interaction API provides plenty of voice potential for developers to experiment with.