ScratchDet

ScratchDet copied to clipboard

对于fasterrcn

1

zhaoyanjoin

Could you provide the weights

1

thx

shengtao96

How to train ScratchDet with mmdetection？

9

thanks for your great repo. I put **ssd300_scratchdet_coco.py** into folder of mmdetection/configs; put **ssd_root34_300.py** into mmdetection/mmdet/model/backbones, and put **ssdbn_head.py** into mmdetection/mmdet/model/anchor_heads. run **./tools/train.py configs/ssd300_scratchdet_coco.py**, there are some errors: **None is...

yanlongbinluck

xjiao004

Hi, I meet some problem using your code. If I use 101 depth of SSDRES512, the output channel number of conv3 is 64, and after some Bottlenecks, the output channel...

herrickli

Hi~I am reproducing your code using pytorch. And I have gained 76.3% with BN in the Backbone. It does improve accuracy. However, I am confused about the BN in the...

LaVieEnRoseSMZ

about some details

4

@KimSoybean HI 有几个疑问： 1. 在分析BN作用时，是基于SSD+VGG16．能不能这么认为，只要是一个包含了BN层的网络，将其作为detector的底层网络时，该detector都可以train from scratch？ 2. 4.1节指出：remove the L2 Normalization．是不是因为BN层的存在可以代替这个L2 Normalization？

foralliance

Will you or did you try Group Normalization

1

According to your paper, BN plays a critical role training from scratch. Did you try Group Normalization ? Does Group Normalization have better perfomance than some pretrained ones. According your...

XCRobert

How do I train my data set?

1

@KimSoybean Thank you very much for sharing the code Can you give me more details?

anjingxing

@foralliance Hi！I think we should speak English because your questions may help the people in other countries.

3

@foralliance Hi！I think we should speak English because your questions may help the people in other countries. 1. No. When you train the model with large input image size (e.g.,...

defei-coder