Spatial AI takes cameras’ understanding to the next level
Spatial AI makes cameras understand and reason about the structure of the surrounding environment in a human-like fashion.
Digital cameras are everywhere. They monitor traffic, assist drivers in cars, and make it possible to stream and post image content on social media. The cameras themselves have come a long way in 20 years, from simple devices that mimic old-school film cameras to intelligent processors that understand images like the visual cortex in the brain. However, they still see images as flat, 2-dimensional, pictures.
Spatial AI makes pictures 3-dimensional. It distinguishes static structures like walls and furniture from people, vehicles and other moving objects. Instead of just knowing where things are in the picture, it knows where they are located in the world. With spatial AI technology cameras can understand their environments like us humans do.
In the intersection of real-time computer vision, sensor fusion and AI
Spectacular AI is a research-based start-up company founded in 2021. It focuses on Spatial AI, a field in the intersection of real-time computer vision, sensor fusion and artificial intelligence.
“Our current product is a software development kit (SDK) that provides visual-inertial SLAM, 3D reconstruction (structure-from-motion), and visual positioning capabilities for various embedded devices. In practice, the SDK produces information about the relative movement of the device, simultaneously with building a map of its environment in real-time”, explains Otto Seiskari, CEO of Spectacular AI.
Optionally, the system can also fuse in absolute position information from GPS or other sources, in which case it can compute an accurate position and orientation in global coordinates.
Technology of future
Spatial AI is a basic building block in quite different applications and devices. It is the foundation of Augmented and Virtual Reality, and the so-called Metaverse.
“We are working with various device hardware manufacturers and aiding them in adding such capabilities to their products—the smart camera devices of the future. Our main target use cases are Augmented and Virtual reality experiences, as well as the autonomous navigation of drones and robots, which, from the technical perspective, are a lot more similar than one might initially think”, says Seiskari.
“It’s not clear when and how exactly these new technologies will start to transform our digital lives, but I would not be surprised if, in a few years, smart phones look as old-fashioned as film cameras look today”, says Otto Seiskari.
Company website: https://www.spectacularai.com/
Contact: otto.seiskari@spectacularai.com
3D city model used in the car video: © City of Espoo - https://kartat.espoo.fi/avoindata/ (https://creativecommons.org/licenses/by/4.0/deed.en)
Spectacular AI is a university spin-off company from Aalto University and Tampere University, founded in 2021.
Read more about research behind Spectacular AI solutions:
O. Seiskari, P. Rantalankila, J. Kannala, J. Ylilammi, E. Rahtu, A. Solin
HybVIO: Pushing the Limits of Real-time Visual-inertial Odometry
2022 IEEE Winter Conference on Applications of Computer Vision (WACV)
A. Solin, S. Cortés, E. Rahtu, J. Kannala
Inertial Odometry on Handheld Smartphones
2018 21st International Conference on Information Fusion (FUSION)
A. Solin, S. Cortés, E. Rahtu, J. Kannala
PIVO: Probabilistic Inertial-visual Odometry for Occlusion-robust Navigation
2018 IEEE Winter Conference on Applications of Computer Vision (WACV)
S. Cortés, A. Solin, E. Rahtu, J. Kannala
ADVIO: An Authentic Dataset for Visual-Inertial Odometry
2018 European Conference on Computer Vision (ECCV)