Megvii-BaseDetection/DeFCN
PyTorch implementation of an end-to-end object detection model using fully convolutional networks.

Velocity · 7d
+0.2
★ / day
Trend
→steady
star history
This repository provides a PyTorch reimplementation of the paper End-to-End Object Detection with Fully Convolutional Network. The project builds on the cvpods framework and implements a detection model trained on COCO, using one-to-many label assignment with NMS-free detection head. It supports multi-GPU and multi-node training via distributed training commands.