xphyro / image-entropy

An ongoing research project on image entropy assessment using machine learning.

image-entropy

This repository contains the source code of an ongoing research project on image/video entropy assessment.

Under src/1-assessment/, the source code of the first phase of this research is contained; namely, comparing and developing different image entropy assessment methods and testing & benchmarking. Requires Python >=3.10.

Under src/2-segmentation/, the chosen entropy assessment method is integrated into a convolutional neural network. Requires Python 3.7.

Under src/3-recreation/, a decoder for src/development/'s output is built. Requires Python >=3.6.

Under src/4-cnn/, CNN model tests are conducted. Requires Python 3.10.

Under src/5-random/, random number generation tests are conducted. Requires Python >=3.6.

Under src/6-video/, 1-assessment/ and 2-segmentation are applied to video. Requires Python >=3.6.

All code are tested on Linux with an NVIDIA GPU. They may not work on other systems.

License

Unless otherwise specified, all code in this repository are licensed under the MIT license. See the LICENSE file for details.