Overview
The rapid rise of artificial intelligence has introduced a wave of new terminology. As a result, people often confuse two concepts: Generative AI (GenAI) and Generalized Scene Reconstruction (GSR). Although both fall under the Spatial AI tree, they serve fundamentally different purposes. GSR measures and creates high-accuracy models of real objects and scenes, while GenAI duplicates patterns to generate fantastical or realistic-looking scenes that are not real.
For a deeper dive, this video breaks down GenAI and GSR with clear examples, including scenarios where the two can work together:
Spatial AI Tree
Our platform, Quidient Reality®, incorporates important functionalities of the Spatial AI tree shown in Figure 1. The primary functions of Quidient Reality® fall under the Generalized Scene Reconstruction (GSR) branch, allowing the engine to reconstruct scenes. The secondary functions fall under the Scene Understanding branch, enabling the engine to segment and label scene parts. The tertiary functions fall under the Generative Spatial AI branch, allowing the engine to complete (at the operator’s option) unobserved parts of the scene.
Figure 1. Spatial AI Tree
Conclusion
As Spatial AI continues to evolve, developing a clear understanding of this emerging terminology is essential. Our goal is to provide valuable, practical insights that help you navigate and make sense of this rapidly changing spatial landscape.
