RTM3D: Real-time Monocular 3D Detection from Object Keypoints for Autonomous Driving
RTM3D: Real-time Monocular 3D Detection from Object Keypoints for Autonomous Driving
I implemented the paper that published at ECCV 2020 from scratch.
Source code: https://github.com/maudzung/RTM3D
Source paper: https://arxiv.org/pdf/2001.03343.pdf
Realtime 3D object detection based on a monocular RGB image
ResNet-based Keypoint Feature Pyramid Network (KFPN) (Using by setting --arch fpn_resnet_18)