Skip to content
IFAI

AI tools

Google Maps rolls out 3D immersive view powered by NeRF

Published Reach 46

Google just rolled out Immersive View on Google Maps, launching first in London, Los Angeles, New York, San Francisco, and Tokyo. The feature uses neural radiance fields (NeRF) and computer vision to turn billions of photos into interactive 3D digital worlds, letting users check venue vibes, weather, and traffic before visiting. AR indoor navigation is also expanding to over 1,000 new transit hubs.

Google Maps rolls out 3D immersive view powered by NeRF
Image: blog.google

Planning a weekend trip usually starts with staring at flat 2D maps and squinting at outdated restaurant photos. Google wants to replace that routine with something closer to flying over a city like a bird, checking out the golden-hour light and restaurant vibes before you even book a flight.

At a live product event in Paris, Google unveiled a massive AI-driven overhaul for Google Maps. The announcements came during a chaotic week for the company, just days after its new chatbot Bard stumbled during a demo and dragged Alphabet shares down over 8%. But tucked beneath the conversational AI noise, Maps quietly rolled out some of its most ambitious computer vision tech to date.

Flying through cities with NeRF

The marquee update is Immersive View. Instead of stitching together static panoramic images, Google is using neural radiance fields, or NeRF, an advanced AI technique that turns ordinary 2D street view and aerial photos into full-fledged 3D environments.

The system reconstructs buildings with realistic materials, shadows, and lighting. You can virtually glide over landmarks like the Rijksmuseum in Amsterdam, spot the entrances, and pull a time slider to see how the weather and crowds shift throughout the day.

Drop down to street level, and you can even peek inside certain cafes and bars to gauge the lighting and vibe. The feature is already live in London, Los Angeles, New York, San Francisco, and Tokyo, with cities like Amsterdam, Florence, and Venice rolling out in the coming months.

Indoor AR arrows and smarter EV trips

Beyond scenic landmark tours, Google tackled everyday navigational friction across transit hubs, walking routes, and electric vehicles.

Indoor Live View, which overlays augmented reality directional arrows directly onto your phone screen via the camera, is embarking on its biggest expansion yet. In the coming months, it will arrive in over 1,000 transit hubs, airports, and shopping centers across cities like Singapore, Taipei, London, and Melbourne, making it vastly simpler to track down a gate or taxi stand.

For electric car drivers using vehicles with Google built-in, Maps will now automatically insert optimal charging stops on routes based on battery levels and real-time traffic. A new very-fast charging filter highlights stations equipped with 150 kW plugs or higher, letting drivers top up in under 40 minutes.

For walkers and transit riders, glanceable directions will push step-by-step turns and dynamic ETAs right onto the lock screen on both Android and iOS 16.1 Live Activities, without forcing users into full turn-by-turn navigation mode.

Why it matters

While rivals and search engines obsess over conversational text boxes, Google is showing what happens when spatial AI meets a massive proprietary data trench. If you travel frequently, playing with Immersive View and indoor AR arrows is well worth your time right now.

We think this is a brilliant monetization of Google's decade-long monopoly on aerial and street imagery. No competitor, including Apple, has the sheer volume of real-world photography needed to train NeRF models at this fidelity.

The limitation, of course, is scale. Rendering high-fidelity volumetric scenes requires hefty cloud computing and crisp local data, meaning these features will remain confined to marquee tourist capitals for quite some time. It is a stunning technology showcase that raises the bar for digital cartography, but universal everyday utility is still years away.

Sources (2)

  1. [1]
    blog.google(opens in a new tab)Originalblog.google
  2. [2]
    siliconangle.com(opens in a new tab)Reportsiliconangle.com

Best of AI, daily