In an exciting development in the world of generative AI, a startup named Odyssey is working on a novel approach that enables creators to develop cinematic worlds with nuanced control over diverse elements including scenery, characters, lighting, and movement. The unique aspect of Odyssey’s technology involves equipping individuals with a sophisticated camera system strapped to their backs, allowing for data collection during hikes in varied terrains around the globe.
Odyssey was founded by Oliver Cameron and Jeff Hawke, former leaders at self-driving technology firms, including Cruise. The company has engineered what it calls an “advanced camera capture system,” which is designed to collect environmental data from almost any accessible location. The apparatus weighs roughly 25 pounds and is equipped with six cameras, two lidar sensors, and an inertial measurement unit, resembling Google’s Street View Trekker.
The camera system possesses the capability to capture the surroundings in “3.5K resolution” and 360 degrees, along with generating “physics-accurate” depth information. This rich capture will allow the company to generate precise digital recreations of real-world settings, akin to Meta’s Hyperscape initiative, covering various landscapes such as forests, caves, and urban environments.
While Odyssey’s ambitious plans seem promising, the efficacy of these reconstructions in improving generative tools for creators remains a focal question. Cameron and Hawke revealed that Odyssey has created several generative AI models, capable of integrating layers of visual detail—including object geometry, lighting, and movement—into a cohesive virtual world conducive to film and game production. Nonetheless, existing “world models” have their limitations, and while Odyssey doesn’t propose to have resolved all issues, the promising technology has garnered substantial backing.
Recently, Odyssey announced it had raised $18 million in a Series A funding round led by EQT Ventures, with additional participation from GV and Air Street Capital, lifting the total investment in the company to $27 million. The influx of funds will primarily support the scaling up of Odyssey’s data collection operations in California, with plans for future expansion to other states and countries, although the startup acknowledges the need for privacy protections in its operations—a lesson learned from Google’s privacy controversies during Street View’s data collection.
Odyssey articulates its vision in a blog post, stating, “We think it will be impossible for generative models to generate Hollywood-grade worlds that feel alive without training on a vast volume of rich, multimodal real-world 3D data.” The company expresses its belief that its advanced generative world-building model will revolutionize content creation across the film and gaming industries, unlocking new potential for storytellers and creators alike.
As Odyssey progresses into wider data collection ambitions, this innovative approach could not only reshape the future of generative tools but also elevate the quality of virtual worlds created in cinematic and game environments, pushing the boundaries of what is possible in digital storytelling. The intersection of AI and real-world data captures exciting prospects for future content creators, as they strive to construct immersive experiences in film, games, and beyond.