Skip to content

Convolutional network for image resolution enhancing based on Tensorflow framework.

Notifications You must be signed in to change notification settings

datbui/enhancer

Repository files navigation

Enhancer

Super resolution convolutional neural network(SRCNN) based on Tensorflow framework.

Prerequisites

  • Python 3.X.X
  • Tensorflow >=1.4.0
  • Scipy>=1.0.0
  • Pillow >=4.3.0
  • Pyyaml >=3.12
  • Numpy >=1.13.3

Instruction

  1. Install Python 3
  2. Follow the official installation process to install TensorFlow(you are supposed to use virtualenv at ~/tensotflow): https://www.tensorflow.org/install/
  3. Install python packages: pip3 install -r requirements.txt
  4. Images should be located in data folder as follows ./data/{dataset}/{subset}/.{extension} (e.g. ./data/cars/train/.jpg)
  5. Preprocess images by preparing tfrecord files: ./scripts/create-tfrecords.sh
  6. Run training ./scripts/start-training-local.sh
  7. TensorBoard is available. Run from commandline: tensorboard --logdir=./summaries/{dataset}/{subset}/logs/
  8. Run prediction ./scripts/start-testing-local.sh

Project structure

  • config.py - configuration script
  • download.py - script to download image sets
  • tfrecords.py - script to create tfrecords
  • model.py - convolutional neural network model
  • main.py - entry point

Sample

Banana
orig
Surface of vinyl disc
orig
Velcro
orig

References

About

Convolutional network for image resolution enhancing based on Tensorflow framework.

Resources

Stars

Watchers

Forks

Packages

No packages published