A lot of home security cameras nowadays have facial recognition, which lets you create a database of friends and family members who regularly visit your house. Then, when the camera sees a face, it determines whether or not it's someone in your database of known faces.
The software can be hit-or-miss, based on a variety of factors, from lighting to changing hairstyles, wearing glasses one day but not the next -- and more.
But one thing we know for sure is that this feature is becoming increasingly popular in our devices, not just in home security cameras, but also our and as efficiency tools helping to . As law enforcement becomes more invested in technology, it's already raising serious questions about privacy and civil rights across the board, and bringing .
But let's step back a bit to the consumer realm. Your home is your castle, and the option of having facial recognition devices therein is still a compelling option for those who want to be on the cutting edge of smart home innovation. Let's take a look at the facial recognition cameras we've tested recently, to see which models are the best and to help you determine if one would work for you.
CNET may get a share of revenue from the sale of the products featured on this page.
If we're talking about sheer facial-recognition capabilities, the Nest Hello, the Nest Cam IQ Indoor and the Nest Cam IQ Outdoor (all of which are essentially the same camera), win by far. Of those models, the Nest Hello is my top pick because it's the least expensive of the three and has the most opportunity to give you important information about who's at your front door.
Nest's IQ Indoor can tell you who's already inside your house, but the Hello, as well as the IQ Outdoor Cam, tell you who's outside your house. The Hello doorbell's eye-level location has the best chance of seeing the most visitors, too (although I suppose you could install the $349 IQ Outdoor cam at eye level, if you wanted).
The snag with the Hello, and other face-tracking Nest cams, is that you do have to pay for the facial-recognition feature. That means you have to subscribe to the Nest Aware cloud subscription service. Learn more about Nest Aware.
Still, the Nest Hello is also a pick for best overall video doorbell. So it's a win/win, whether or not you want to enable facial recognition.
The Tend Secure Lynx only costs $60. Given that, I was skeptical that this camera would deliver, but it does. Not only does the camera itself perform well and offer some nice features like free seven-day event-based video clip storage, but it also has facial recognition free of charge (unlike the optional Nest Aware service).
Create your database of familiar faces, and the Lynx takes over. There is a bit of a learning curve as it becomes familiar with each face, but it's a very good option if you want an inexpensive indoor home security camera with decent facial recognition.
The $299 Nest Cam IQ Indoor is very similar to the Nest Hello doorbell. It has facial recognition (if you sign up for a Nest Aware subscription) and lets you know who walks in front of the camera's field of view with consistent accuracy.
But it also has some additional benefits. Because it is an indoor camera, Nest gave it an integrated Google Assistant speaker. That means the camera essentially doubles as a Google Home speaker and can answer basic questions like what the current weather or traffic is in your area -- and control a variety of Google-Assistant-enabled smart home devices. It also works with Amazon Alexa.
Facial recognition cameras: Every one we tested
Here's a recap of the facial recognition cameras we've tested recently.
Worth considering, but not as good as the top picks above:
- - The IQ Outdoor camera is similar to the $229 Nest Hello and the $299 IQ Indoor when it comes to specs and performance, but it offers a worse value at a whopping $349 per camera.
- - Netatmo's Welcome indoor camera did a fair job detecting faces, but the feature ultimately wasn't quite as reliable as we'd like.
- Wisenet SmartCam N1 - The $150 SmartCam N1 did a good job detecting faces and it comes with a built-in microSD card slot for local storage, but the $60 Tend Secure Lynx performs just as well for much less.
- Unreliable performance, including its facial recognition tech, seriously hurts this all-in-one system's appeal.
- Tend Secure Lynx Pro - While the indoor-outdoor Lynx Pro is technically the high-end version of the indoor-only Lynx, its improved specs didn't translate to better facial recognition.
Note that the recommendations above were at the time of testing, and could change based on later software updates. We'll periodically update this list as such changes warrant.
How we tested
When setting up a camera with a facial-recognition function, you create profiles of individual people, by either taking their picture in real time and adding it, or using an existing photo that you have of them. From there, The camera should be able to distinguish human faces from every other type of motion activity and single out the ones it recognizes from your database of familiar faces. When it's working optimally, you will get an alert that says the camera saw "Chris," "Molly" or whoever is in your database.
There are many use cases for this type of functionality, but some common ones include getting an alert when your kids get home from school, or if a dog walker or a family caregiver shows up. It creates peace of mind when you're expecting someone to show up and you want an automated alert telling you they have (especially when you aren't home to greet them).
But it also helps in security scenarios, since the camera is essentially distinguishing between faces it recognizes and those it doesn't. That way, if your camera sends you an alert that it saw someone on your front porch or walking into your house, but you don't recognize them, you can more quickly send the information to law enforcement in the event of an actual break-in or theft, instead of having to sift through dozens of generic motion alerts to find the activity.
The best way to test these cameras is to create a database, which is what I do when I test a camera with facial recognition (see the screenshots above). I add people to my database and let the camera do the rest. It's best to give these cameras at least a few days, because some improve significantly, even over a short period of time, as they see faces at different angles.
Then it's a matter of looking at how well the camera actually recognized faces. How often did it correctly identify my face versus someone else's face? How did it do when approached at different angles and changes to hairstyles and clothing accessories? Was the camera able to detect faces at all? Some occasionally struggle to see any faces, even ones that claim to have facial recognition, and instead mark the activity as a basic motion alert (ahem,).
The future of facial recognition
Amazon's doorbell and security camera company, Ring,in 2018. The patents suggest that future Ring products might be able to identify faces from "most wanted" lists and automatically contact law enforcement. Here's an excerpt from one of the patent filings:
A video may be analyzed by an A/V recording and communication device that recorded the video (and/or by one or more backend servers) to determine whether the video contains a known criminal (e.g., convicted felon, sex offender, person on a "most wanted" list, etc.) or a suspicious person. Some of the present embodiments may automatically submit such a video stream to the law enforcement agencies.
"Amazon is dreaming of a dangerous future," ACLU attorney Jacob Snow said in a blog post.
"The history of discriminatory government surveillance makes clear that face surveillance will disproportionately harm people already targeted by the government and subjected to racial profiling and abuse — immigrants, people of color, and the formerly incarcerated," Snow added.
Right now, Ring cameras don't offer facial recognition at all. Models that do, like the Nest Hello, are only designed to ID faces you add to your list of "familiar faces." They won't draw from a law enforcement list to determine if a convicted felon is nearby -- or reach out to law enforcement if they spot a face that could match someone in a database.
While we know of no ethical breaches associated with these cameras on the market right now, the reality is we have no way to verify how the data is used. Even if we give the companies involved the benefit of the doubt regarding their data usage policies, those policies could change at any time. And when you consider that Ring is owned by Amazon and Nest is owned by Google, the potential for a Big Brother scenario is readily apparent.
We'll continue to keep an eye on home security cameras, doorbells and other devices with built-in facial recognition tech, to follow along with any changes in industry trends -- and to see if any new models come close to matching the smarts of Nest's Hello buzzer.