Xilinx is now part ofAMDUpdated Privacy Policy

Real-time Multi-Task ML Perception Demo for Automotive

Xilinx multi-task model aims to do different image perception tasks simultaneously while achieving high performance and efficiency. With the optimized MTLv3 model and powered by Vitis AI, now there are five tasks for vehicle detection, lane detection, segmentation, drivable area detection and depth estimation achieved based one model and running at real-time speed on edge.