Odyssey, a startup, is working on developing a revolutionary AI system designed to provide content creators with unprecedented cinematic world generation capabilities. Through a unique field data acquisition method, they use teams equipped with advanced camera systems to collect real scene data around the world, providing training materials for AI models. Recently, Odyssey completed a US$18 million Series A financing, with a total financing amount of US$27 million, which will help it further expand its data acquisition business.
A remarkable startup is developing a breakthrough AI system that promises to bring unprecedented cinematic world generation capabilities to creators. The company, called Odyssey, has adopted a unique approach: staff are carrying special camera systems to collect scene data from around the world on the spot.
Founded by senior experts in autonomous driving Oliver Cameron and Jeff Hawke, Odyssey recently completed a $18 million Series A funding round led by EQT Ventures, with GV and Air Street Capital. So far, the company's total financing amount has reached US$27 million. The funding will be primarily used to expand its data acquisition operations in California.
The advanced photography system developed by Odyssey weighs about 25 pounds and is equipped with 6 cameras, 2 lidar sensors and an inertial measurement unit. This system is similar to Google Street View Trekker, which can perform 360-degree panoramic shooting at 3.5K resolution, while recording depth information metadata with physical accuracy.
The company's core technology is to input the collected data into the algorithm to capture the detailed features that constitute the real world. This process is similar to Meta's Hyperscape project, which allows digital reconstruction of real scenes such as forests, caves, trails, beaches, glaciers, parks and buildings. Odyssey has developed multiple generative AI models that create layers of visual detail including object geometry, lighting, and motion and combine them into a single virtual "world" to create the desired scene.
"United models cannot create a viable Hollywood-class world without training based on a large number of rich multimodal real 3D data," the company said in its blog. "We believe that an advanced generative world building model will Bringing revolutionary changes to the creation of content such as movies and games. "
Although there are still certain limitations in the "world model", Odyssey plans to expand the scope of data collection to other states and countries in the future. Given the privacy protection issues (Google Street View team has been under scrutiny by regulators for photographing images of public places that violate the privacy of passers-by), the company needs to take appropriate privacy protection measures during its expansion.
The advent of this technology may revolutionize the way film and television production and game development are being developed, providing creators with more efficient and economical content production solutions. With the continuous advancement of AI technology, we may soon see the application of this revolutionary technology in more fields.
Odyssey's AI system, with its unique field data acquisition and powerful generative models, is expected to completely revolutionize the field of content creation. In the future, as the technology becomes more mature and the application scenarios expand, we will look forward to it bringing more surprising results and pay close attention to its measures in privacy protection.