Recogni provides an integrated module which inclides three passively cooled image sensors and an external depth sensor. The perception is perfromed by an ASIC with 1 peta operations per second
that captures three uncompressed 8-12 MP streams at 60 fps. while consuming about 8 watts of power. One power saving approach is offloading processing to multiple points on a vehicle.
Recogni boldly claims its system outperforms rivals by more than two orders of magnitude on perception tasks like image classification, objection detection, action anticipation, and depth inference. On DepthNet, Recogni says it’s capable of analyzing 3,500 scenes per second.
Recogni intends to target level 2 AVs and plans to pivot to platforms for level 3 and level 4 cars