FGD
FGD copied to clipboard
yzd-v
Focal and Global Knowledge Distillation for Detectors (CVPR 2022)
Hey,bro. I wonder using two stage method like Faster-RCNN to distill one stage method like Yolo, only using neck feature to distill, will it work？what's your opinion? Thanks.
当前有两个YOLOX-l模型（分别检测抽烟（YOLOX-l-cigar）和玩手机（YOLOX-l-cellphone）），利用FGD将YOLOX-l-cigar和YOLOX-l-cellphone模型对YOLOX-m学生进行训练，训练后的模型既能检测抽烟 也能检测玩手机，有相关的示例代码或者指导吗？多谢
When I read the `mmdet/distillation/losses/fgd.py`, there is something I can't understand. 1. In the init function of the `FeatureLoss`, at the end of that function, after all the initialization, why...
I am using *fgd_retina_r101_fpn_2x_distill_retina_r50_fpn_2x_coco.py* to distill Retinanet-r101 **(Using my own CoCo formatted)** with mAP 0.75 but my Retinanet-r50 mAP significantly drops to 0.58. I am using SGD with same parameters...
Have anyone tried to distill Yolox-l into Yolox-s with COCO pretrained weights and get better results? or even smaller student models?
I use retinanet to train and test on VOC dataset. I found that the accuracy of retinanet after training with FGD is even lower than the original retinanet. It seems...
您好，我在跑实验的时候出现了这个错误，想请问一下是什么原因导致的，有可以解决的方法吗？  mmcv和mmdet都用的是和作者相同的配置
# Welcome update to OpenMMLab 2.0 I am Vansin, the technical operator of OpenMMLab. In September of last year, we announced the release of OpenMMLab 2.0 at the World Artificial...
Dear authors, Kindly share the implementation details 1. Pytorch version 2. Cuda version I am getting multiple issues due to missing these information in the paper and code itself.
