Amazon has added voice recognition to Alexa. The new feature allows Alexa to know who in your household is talking and give personalized responses, eliminating the need to switch profiles.
Voice recognition for Alexa is starting to roll out today and should be available on all devices by the end of the week. In order to use the feature, you must first teach Alexa your voice and the voice of anyone else in your household that you want Alexa to recognize.
To teach Alexa your voice, open the Alexa app and select Settings from the menu in the upper left corner of the screen. Scroll down to the Accounts section of the Settings screen and select “Your Voice” from the list of options.
You’ll then be asked to select one of your Alexa devices to use as the voice training tool. You’ll want to select a device that is in a quiet room because you’ll be asked to speak 10 sentences to the Alexa device. It’s a good idea to mute the microphone on any other Alexa devices that might be close enough to hear you speaking to the device you’re using for voice training, since the 10 sentences are all standard Alexa requests.
After completing voice training, it’ll take about 15-20 minutes for your Alexa devices to distinguish your voice from the voice of others. You can ask “Alexa, who am I?” to know when Alexa has learned your voice.
In order to train your Alexa devices to distinguish the voices of other people, you have two options. If the person is a member of your Amazon Household, they can simply log into the Alexa app from any device using their own Amazon login information and complete the same voice training steps that you completed.
If the person is not part of your Amazon Household or they don’t have an Amazon account of their own, you can use your own Amazon account to train their voice. You’ll first need to log out of the Alexa app and then log back in using your own Amazon account. After logging in, you will be asked if you are yourself or if you are someone else. Select that you are someone else and enter the name of the person that you want to teach Alexa to recognize. That person will then go through the voice training steps and will be recognized by your Alexa devices.
Once Alexa is able to distinguish between different people, it will be able to give custom responses to each person, without needing to first switch profiles. Alexa will automatically learn each person’s music tastes based on their voice and what they commonly ask to play. Each person will then be able to simply say “Alexa, play music” or “Alexa, play a station” and receive recommended music that matches their taste.
Alexa’s calling and messaging capabilities will also automatically recognize who is making a request and respond appropriately. When someone that Alexa recognizes asks to call a specific individual, Alexa will know whos list of contacts to use, regardless of which profile is currently the active profile. Asking Alexa to “play my messages” will also automatically play only the messages sent to the recognized individual and saying “send a message” will mark the message from the recognized speaker.
Voice shopping now becomes much more secure with recognized voices. Alexa will no longer need to ask for the speaker’s voice shopping PIN before placing an Amazon order if the speaker’s voice is recognized. Another handy improvement comes to flash briefings. If you ask to play your flash briefing, Alexa will remember which stories you’ve already heard if it recognizes the speaker and will only play stories you haven’t heard yet if you ask for the news again later in the day.
This is probably just the start of how deep voice recognition will be integrated into all the various things Alexa can do. Assuming it works correctly, it should make using Alexa devices in a home with multiple people much more secure and pleasant.