search

BangguWu / ECANet

Code for ECA-Net: Efficient Channel Attention for Deep Convolutional Neural Networks
MIT License
1.24k stars 197 forks source link

Need to change state_dict while using pretrained model? performance not going up #29

Closed hezhu1996 closed 4 years ago

hezhu1996 commented 4 years ago

@BangguWu Very interested in your work which is impressive. I tried to reimplement your work in detectron2（FasterRcnn）but performance doesn't really go up. I notice that you evaluate object detection task using mmdetectrion framework, would you please tell except for adding the ECA Layer into ResNet, is there any other place I should modify?

One more question, when training on COCO dataset, did you use the pretrained (ECA+ResNet) backbone or just pretrained (ResNet) backbone? Is there a large difference when doing detectrion task in COCO? Thanks

hezhu1996 commented 4 years ago

One more question, when training on COCO dataset, did you use the pretrained (ECA+ResNet) backbone or just pretrained (ResNet) backbone? Is there a large difference when doing detectrion task in COCO? Thanks

BangguWu commented 4 years ago

The object detection and instance segmentation experiments we performed using mmdetection only replaced bakcbone, and the specific modification method can be combined with the code we gave and refer to https://github.com/open-mmlab/mmdetection/blob/master/docs/tutorials/new_modules.md for modification.

when training detector, we used the pretrained ECA-Net model , The performance of the backbone is an important factor that affects downstream tasks, because it can learn better representation.