DenseTeacher

DenseTeacher copied to clipboard

1shenhui

Hi, how did you implement the Hard Negative Mining technique mentioned in the paper? cause I have not seen it in the code.

zfli-xdu

Cannot reproduce the results of the paper

3

Hi, thanks for your excellent work and the open-source. I am trying to reproduce the results reported in the paper. Currently, I conduct the experiment of 1% COCO. I only...

kinredon

自有数据集

1

你好，dense teacher支持在自有数据集上训练吗

ZayneYe

I noticed that the batchsize dimension is concatenated together when selecting the top 1% learning region, does this cause the batchsize to have a big impact on performance?

PlutoQyl

install errors

3

Thanks for your interesting work! I have installed cvpods, and import cvpods success, but converted to coco-p10 directory and run "pods_train --dir ." and got error: AttributeError: 'SemiRunner' object has...

XipengY

是否提供模型权重

2

您好！我打算在mmdetection框架下复现dense teacher，请问能提供模型权重做推理精度对齐吗？

shufanwu

请问一下Dense teacher提出的DPL是否其实只是针对预测的类别分支？

2

在图1中，有Dense Pseudo-label 和 Pseudo-box Label两个分支，分别计算了Ldpl损失和Lbox损失，所以Ldpl其实就是对应教师模型输出分类分支后的sigmoid所做的吗？整个模型还是用了NMS来生成伪框监督回归分支的损失计算？ 3.3 Dense Pseudo-Label Since the learning region is selected, unsupervised learning for regression branch can be easily achieved. 3.3中关于回归的分支的叙述较少，上面说由于选择了学习区域，因此回归分支的学习更容易实现，这个更容易实现指的是什么呢？ 希望作者能够解答一下 感谢

joeyslv

[Dense-Teacher implemention !!! ] Dense-Teacher has been implemented in PaddleDetection, welcome to use it for research or projects!

1

[DenseTeacher implemention !!! ] DenseTeacher has been implemented in PaddleDetection, welcome to use it for research or projects! https://github.com/PaddlePaddle/PaddleDetection/tree/develop/configs/semi_det/denseteacher https://github.com/PaddlePaddle/PaddleDetection/tree/release/2.6/configs/semi_det/denseteacher

nemonameless

The result of FCOS used in Unbiased Teacher V1 is not consistent with those in Dense Teachers. There is a big difference between the two results. Can you supply some...

Tourbillon