This article mainly introduces two practical concepts in transfer learning: pretraining and finetuning.

Supposing that you have the experience of developping a CNN using pytorch.

For example in function forward, you write

1
2
3
4
5

h = self.conv1(x)
r1 = self.pool1(h)

h = self.conv2(r1)
r2 = self.pool2(h)

Pretraining means that the 1st layer like conv1 and pool1 is done for you by the pretrained model. And the parameters of them are fixed. What you need to do is just finetuning, i.e. you input the model with x as input but the r1 is fixed so your truly take r1 as input and train the later layers’ parameters.

Usually, the pretrained model is in large scale. To get the inside layers of this network, we use some functions like mx.symbol.FullyConnected to get the fine tuned data.

Histogram equalization is a special case of histogram matching.

Read More

Some of the essences of neural networks.

Read More

The discrict 2D cross correlation is very similar to 2D convolution.

Read More

VQA(Video quality assesment) is generated from image quality assessment.

Read More

The stochastic gradient descent method is commonly used in machine learning.

Read More

Pandas is a wide-spread used data structure in Tensorflow, the famous architecture in Machine Learning.

Read More

Notes includes Windows subsystem Linux, VMWare Virtual machine, remote cloud and dual system.

Read More