Enabling Indoor Augmented Intelligence by Recognizing the world


Modeling indoor environments from videos

We use computer vision techniques to build 3D models of indoor environments. The input can be videos took by smart phones. The 3D model can be further used for building floor plan, provide localization and navigation services.

We are collaborating with CubiCasa to convert video recording into floor plan ( VimAI is responsible for 3D modeling.

Indoor scene understanding

We apply machine learning techniques to analyze the collected videos to recognize objects and undertand scenes. The recognized contents can be accurately mapped to the 3D spcae. 

Indoor localization and navigation

We can provide high accurate indoor localization and navigation services. The user only needs to simply take a photo. VimAI will be responsible to calcuate where the user locates and help the user to navigate to the point of interest.

Augmented reality 

With the advances of 3D modeling, indoor localizaiton and navigation, and scene understanding, we can further provide augmented reality services. Customers who want to create "Indoor Pokémon Go" can benifit from this technique.
  1. Find a coffee machine in an office building
  2. Find products in a grocery store
  3. Product detection in a supermarket