https://github.com/mountchicken/neural-style-transfer

New Year, New Style For Your Image
https://github.com/mountchicken/neural-style-transfer

pytorch styletransfer

Last synced: 18 days ago
JSON representation

New Year, New Style For Your Image

• Host: GitHub
• URL: https://github.com/mountchicken/neural-style-transfer
• Owner: Mountchicken
• Created: 2021-02-17T07:53:41.000Z (over 3 years ago)
• Default Branch: main
• Last Pushed: 2021-03-31T12:37:23.000Z (over 3 years ago)
• Last Synced: 2024-10-06T11:20:54.588Z (about 1 month ago)
• Topics: pytorch, styletransfer
• Language: Python
• Homepage:
• Size: 1.34 MB
• Stars: 0
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
# Neural-Style-Transfer

New Year, New Style For Your Image

- ![test_example](https://github.com/Mountchicken/Neural-Style-Transfer/blob/main/github/image1.png)

- ![test_example](https://github.com/Mountchicken/Neural-Style-Transfer/blob/main/github/sky.jpg)

- - ![test_example](https://github.com/Mountchicken/Neural-Style-Transfer/blob/main/github/ddd.png)

## Structure

### File

- `main.py`: models and codes for NST,run it

- `content`: where you stored the content image

- `style`  : where you stored the style image

- `generated`:The genearted images

## Which hyperparameters Can be changed in the main.py

### line33--image_size

- this parameter determines the size of the geneareted image and the size which the content image will

- be resize to. The larger it is, the better the quality of the geneareted image will be with the cost

- of more training time

### line45-line46 --steps and lr

- you can use a decreasing learning-rate and less or more steps depends on the training result

### line47-line48 --alpha and beta

- The larger the alpha, the more the generated image looks like the content iamge.The larger the beta,

- the more the generated image looks like the style image

## contact information

- [email protected]