This generative AI startup is attaching cameras to people’s backs

A startup is training an AI system that, it claims, one day will enable creators to envision and realize cinematic worlds from scratch, controlling scenery, characters, lighting, and motion.
This generative AI startup is attaching cameras to people’s backs

A startup is training an AI system that, it claims, one day will enable creators to envision and realize cinematic worlds from scratch, controlling scenery, characters, lighting, and motion. How? By sticking cameras to human backs and sending them on hikes round the world.

Odyssey, founded by self-driving pioneers Oliver Cameron and Jeff Hawke (Cameron was formerly the VP of product at Cruise), claims to have developed an "advanced camera capture system" that can capture data nearly anywhere a person can go. Weighing in at about 25 pounds, the system packs six cameras, two lidar sensors, and an inertial measurement unit.

It is akin to Google's Street View Trekker in that it captures the surroundings in "3.5K resolution" and 360 degrees, with "physics-accurate" depth information metadata attached.

What is the end of that? Well, Odyssey says it is taking data from the system and feeding it through algorithms to "capture fine details making up our world." In essence, the company is creating digital reconstructions of real-world scenes à la Meta's Hyperscape project - scenes with forests, caves, trails, beaches, glaciers, parks, buildings, and so forth.

Not clear is how exactly these reconstructions will make their way to better generative tools for creatives. Cameron and Hawke have said over time that Odyssey has developed a number of generative AI models creating layers of visual detail from object geometry to lighting to motion. Then, all of these stitch together into a single virtual "world" to construct desired scenes.

Even the best "world models" today have limitations, however — and Odyssey does not claim to have solved all these. Nevertheless, it's securing cash to forge ahead.

Odyssey said today that it had raised $18 million in a Series A funding round led by EQT Ventures and including GV and Air Street Capital in the round. With the new funds, which puts total capital raised at $27 million, the company will expand its data collection in California operations.

The company plans to expand data collection into other states and countries in the future-one hopes, with protections for privacy. Google's Street View team, for example, has found itself in regulators' crosshairs for snapping images of public places that invade the privacy of bystanders.

"We believe it will be impossible to develop Hollywood-grade worlds that feel alive without training on a vast volume of rich, multimodal real-world 3D data," the company wrote in a post on its blog. "We think an advanced generative world-building model will unlock a better way to create film, games and much more."

Blog
|
2024-11-14 18:32:33