Activations of freezed layers are different between before/after training, why?

1 次查看（过去 30 天）
ntinoson
ntinoson 2018-6-29
评论： Amanjit Dulai 2018-8-28
I follow the example "transfer-learning-using-googlenet" where, the last 3 layers ('loss3-classifier','prob','output') are replaced with 3 new ones. Then I 'freeze' the first 141 layers (that is up to and including 'pool5-drop_7x7_s1'):
layers(1:141) = freezeWeights(layers(1:141));
lgraph = createLgraphUsingConnections(layers,connections);
Then I follow fine-tuning.
Since 'pool5-7x7_s1' is BEFORE 'pool5-drop_7x7_s1', I would expect that the following two vectors were the same:
b_orig= activations(net_orig, I, 'pool5-7x7_s1');
b_tune= activations(net_tune, I, 'pool5-7x7_s1');
but they aren't!... Any idea why?
p.s. I also tried the activation of several other layers BEFORE 'pool5-drop_7x7_s1', and I got different vectors.... 'I' is an image, 'net_orig=googlenet;', and 'net_tune' is the resulting net after tuning.
  2 个评论
ntinoson
ntinoson 2018-7-17
I did also try other CNNs, and same result. Activation of freezed layers after fine-tuning is different to that before fine-tuning (for the same input image of course). If anyone comes up with an explanation, drop a line!

请先登录，再进行评论。

请先登录，再回答此问题。

采纳的回答

Amanjit Dulai
Amanjit Dulai 2018-8-14
The vectors are different because when you fine tune on a new dataset, the average image in "imageInputLayer" is recalculated for your new dataset.

更多回答（0 个）

类别

在 Help Center 和 File Exchange 中查找有关 Image Data Workflows 的更多信息

标签

产品


版本

R2018a

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by