First announced in May 2017, the security camera created by Lighthouse AI has been officially launched. The product attracted the market’s attention for its deep learning and 3D-sensing technology which is claimed to be able to distinguish humans and other things in the environment.
First announced in May 2017, the security camera created by Lighthouse AI has been officially launched. The product attracted the market’s attention for its deep learning and 3D-sensing technology which is claimed to be able to distinguish humans and other things in the environment.
Lighthouse identifies different objects with a RGB camera and a 3D sensor. By recognizing more patterns and objects over time, the deep learning technology can better distinguish various types of objects and tell whether it’s a family member or not.
Up until now, however, object recognition or facial recognition is not new in the industry. Cameras such as Nest Cam IQ or Netatmo Welcome can learn human faces as well. What makes Lighthouse stand out is how the AI technology presents findings to users with natural languages and voice.
“We wrestled for a long time with what is the right mix of buttons and UI features. It turns out it’s just dramatically better to just have people ask,” said Alexa Teichman, CEO and co-founder of Lighthouse AI, in an interview with The Verge.
In the Lighthouse app, users can simply ask the security camera questions by pressing the “ask” tab. Native questions and commands like “Did you see any new people today?” “What did the kids do while I was out today?” and “When did the dog walker come here today?” can be understood by the app which will then provide answers by showing related video clips.
Users can also set triggers using natural language or voice in the app. They may say “tell me if you don’t see someone enter the home between 3pm to 5pm” or “tell me when my kids get home.” Therefore, notifications for certain events or absent activities can be set up as well.
Its AI and command interface prevents users from scrolling video recording every day if they want to know what happened in the house. No fixed terms of commands are required to talk to the Lighthouse, making the interaction more natural, better than integrations with Alexa or Google Assistant.
The company will continuously roll out new features via updates. For now, it can tell when kids are running and hand-waving is under beta testing, which allow the kids to get their away parents’ attention simply by waving to the camera.
Lighthouse camera includes a RGB camera, 3D sensor, speaker, microphone and siren. The camera captures 1080p HD video and features night vision. The company said user data is fully protected with bank-level security system. Users’ video footage will be stored for 30 days and then deleted permanently.
Lighthouse AI has raised US$17 million to date and is part of Playground incubator, which is owned by Andy Rubin, the creator of Android system. Both of its co-founders have backgrounds in self-driving car, where 3D-sensing is utilized to see the structure of the environment and to determine the location of objects. The Lighthouse’s background makes the market believe the startup does have the technology it claims to have.
The product is on sale now for US$299, while its service fee costs US$10 per month.