One of the startups active in the field of artificial ielligence recely claimed that it was to create a tool to create a cinematic world similar to the real world. To achieve its goal, this startup attaches a 360-degree camera to people to provide the necessary data to train its models.
Generative AI can generate a variety of images, text, and even music, but what if we used it to create virtual worlds that closely resembled the real world? In this regard, the newly founded startup Odyssey iends to improve the use of generative models by one step and provide developers with the possibility of creating cinematic worlds. Of course, to achieve such a goal, not only an advanced software model is needed, but also special cameras are used to collect data.
Odyssey startup has announced that it has developed special cameras with an advanced image recording mechanism that will be used to collect data from differe environmes. Each of these cameras weighs about 11 kg and is placed on people’s shoulders in the form of a backpack. Among their features, we can meion 6 differe lenses and 2 lidar sensors, which are placed next to an inertial measureme device (IMU) with a combination of accelerometers and gyroscopes.
What is the use of creating a supernatural virtual world?
The camera iroduced by Odyssey is very similar to Google’s cameras, which are used to record images in the Street View mode of Google Maps. Odyssey managers claim that their camera can perform 360-degree imaging with 3.5K resolution and add additional data such as the depth and distance of objects from the observer to the output file. Finally, the obtained information is processed by special algorithms to be used for the developme of productive artificial ielligence with the possibility of building virtual worlds.
Odysse is not the first company that thought of building virtual worlds; Meta has a similar project called Hyperscape. Although it is not yet clear how the process of building such worlds will be done, Odyssey’s founders meion that multiple models will be used to design the geometric shapes, lighting and motion engine, and their output will eveually be merged together. The company has managed to raise $27 million in capital so far, but it will probably need more money to achieve the desired result. Game industry and film studios will be poteial customers of such a product.
