Pokemon GO developer Niantic has announced on your blog that she was busy building a Large Geospatial Model (LGM) to help the company achieve spatial intelligence. Game developer he says that Niantic’s large geospatial model, which uses machine learning to “understand a scene and connect it to millions of other scenes globally.” However, the data for this language model seems to have come from unwitting Pokemon GO players around the world. Those who haven’t read the terms and conditions of the mobile game won’t know about the geospatial technology and player recordings that are on by default and can be turned off. LGM is built on the studio’s Visual Positioning System and aims to allow Niantic to “implement a shared understanding of geographic locations and understanding of places that have yet to be fully scanned” that companies could potentially use.
“Similar to the Large Language Model (LLM), Niantic’s model draws data from real-world locations and hopes to use technology to “enable computers not only to perceive and understand physical spaces, but also to interact with them in new ways. .” And what makes its data more meaningful than something like Google Maps or Street is the point of view: As Niantic notes, the data is “taken from a pedestrian’s perspective and includes places inaccessible to cars.” – Game Developer.com