Products on Show
Synthetic data for training and testing
Anyverse - Booth: AV222
One of the major disadvantages of using real-world data for the training and testing of self-driving vehicles is that it is practically impossible to record and process all possible scenarios and, even then, the quality is not always guaranteed. Anyverse provides virtual, synthetic data that is physically correct and as close to reality as possible. These attributes contrast with often-used data from game engines, which may look nice, but is often not physics-accurate. What’s more, Anyverse’s data is highly scalable and comes complete with metadata including lighting, sensor and camera specifications. Clients can have fine control over data set production and variability, counting on maximum similarity with reality.
ANYVERSE
Booth: AV222