Skip to content

Commit

Permalink
Update README.md (#43)
Browse files Browse the repository at this point in the history
* Update README.md

Added additional requirements

* Update README.md

* Update README.md

* Update README.md

* Update README.md

* Update main.py

Added fix from Issue titu1994/Image-Super-Resolution#27 (comment)

* Update models.py

Closer to working than the original code (still doesn't work)

* Update img_utils.py

Closer to working than the original code (Still doesn't work)

* Update advanced.py

Closer to working than the original code (Still doesn't work)

* Delete advanced.py

* Delete img_utils.py

* Delete main.py

* Create advanced.py

* Create img-utils

* Create main.py

* Update models.py
  • Loading branch information
CaptainStabs authored and Erman Nghonda committed Aug 17, 2021
1 parent 4d86626 commit db88b51
Show file tree
Hide file tree
Showing 3 changed files with 1 addition and 4 deletions.
1 change: 1 addition & 0 deletions README.md
Original file line number Diff line number Diff line change
Expand Up @@ -8,6 +8,7 @@ Also contains models that outperforms the above mentioned model, termed Expanded
## Setup
Supports Keras with Theano and Tensorflow backend. Due to recent report that Theano will no longer be updated, Tensorflow is the default backend for this project now.

Requires Pillow, imageio, sklearn, scipy, keras 2.3.1, tensorflow 1.15.0
## Usage

**Note**: The project is going to be reworked. Therefore please refer to [Framework-Updates.md](https://github.com/titu1994/Image-Super-Resolution/blob/master/Framework-Update.md) to see the changes which will affect performance.
Expand Down
1 change: 0 additions & 1 deletion img_utils.py → img-utils
Original file line number Diff line number Diff line change
Expand Up @@ -12,7 +12,6 @@

'''
_image_scale_multiplier is a special variable which is used to alter image size.
The default image size is 32x32. If a true upscaling model is used, then the input image size is 16x16,
which not offer adequate training samples.
'''
Expand Down
3 changes: 0 additions & 3 deletions models.py
Original file line number Diff line number Diff line change
Expand Up @@ -31,10 +31,8 @@
def PSNRLoss(y_true, y_pred):
"""
PSNR is Peek Signal to Noise Ratio, which is similar to mean squared error.
It can be calculated as
PSNR = 20 * log10(MAXp) - 10 * log10(MSE)
When providing an unscaled input, MAXp = 255. Therefore 20 * log10(255)== 48.1308036087.
However, since we are scaling our input, MAXp = 1. Therefore 20 * log10(1) = 0.
Thus we remove that component completely and only compute the remaining MSE component.
Expand Down Expand Up @@ -134,7 +132,6 @@ def upscale(self, img_path, save_intermediate=False, return_image=False, suffix=
patch_size=8, mode="patch", verbose=True):
"""
Standard method to upscale an image.
:param img_path: path to the image
:param save_intermediate: saves the intermediate upscaled image (bilinear upscale)
:param return_image: returns a image of shape (height, width, channels).
Expand Down

0 comments on commit db88b51

Please sign in to comment.