使用 Tensorflow 2 对象检测 API 冻结了哪些层?

Which layers are frozen using Tensorflow 2 Object detection API?

我如何了解哪些层被冻结以微调来自 Tensorflow Model Zoo 2 的检测模型? 我已经成功设置了 fine_tune_checkpoint 和 fine_tune_checkpoint_type: detection 的路径,并且在文件原型中我已经读到“检测”意味着

// 2. "detection": Restores the entire feature extractor.
The only parts of the full detection model that are not restored are the box and class prediction heads. 
This option is typically used when you want to use a pre-trained detection model 
and train on a new dataset or task which requires different box and class prediction heads.

我不太明白那是什么意思。在这种情况下,恢复意味着冻结?

据我了解,目前 Tensorflow 2 对象检测在从微调检查点进行训练时不会冻结任何层。报告了一个问题 here to support specifying which layers to freeze in the pipeline config. If you look at the training step function,您可以看到在训练期间应用梯度时使用了所有可训练变量。

此处恢复意味着模型权重是从检查点复制的,用作训练的起点。 Frozen 意味着在训练期间权重没有改变(即没有应用梯度)。