YLOMF: You Look Once for Multi-Feature—A Multi-Feature Multi-Task Network for Intelligent Real-Time Perception in Autonomous Driving by B. N. Anoop, S. K. Abhilash, Venu Madhav Nookala, S. Prita, S. Raghavendra

“…It achieves 85.1% IoU for lane segmentation, 81.56% accuracy in 2D object detection, and 31.5% (easy), 26.38% (medium), and 23.9% (hard) accuracy in 3D instance detection, surpassing state-of-the-art benchmarks. …”
Get full text