Skip to content

Latest commit

 

History

History
79 lines (55 loc) · 1.85 KB

README.md

File metadata and controls

79 lines (55 loc) · 1.85 KB
Raw

Computer Vision Project: Medical Image Forgery Detection

All deliverables (codes, reports) for CVI Semester Project - Fall 2021

Priscila Moreira ([email protected]) and Mahsa Mitcheff ([email protected])

Deliverables

Report 1

Report 2

Final Report

Final Presentation

Short video on SSD performance

Instructions to run our trained SSD model on test set images

Set the envirement

1. Create a conda environment using the following command: Note: only on GPU

conda create -n py368-tf115 python==3.6.8 tensorflow-gpu==1.15.0 cudatoolkit=10.0

2. Activate the created envirement

Using conda 4.4:

conda activate py368-tf115

or using conda versions older than 4.4:

source activate py368-tf115

3. Install opencv

conda install opencv

4. Install matplotlib==3.3.2

pip3 install matplotlib==3.3.2

5. Install jupyter notebbok to run the script

conda install jupyter

Download our repo and demo folder inside of CVI-project

cd CVI-project
wget http://www.crc.nd.edu/~pmoreira/CVI-project/demo.zip
unzip demo.zip

Inside the demo folder you shoul see

  • ckeckpoints/ - folder with graph and weights of the proposed model
  • test_sample - foler with some images from our test set (unseen data during the trainig)
  • test_solution.ipynb - the jupyter notebook for testing the trained model

Run the code using jupyter notebook

Inside the folder ./demo, run the script test_solution.ipynb.

jupyter notebook ./demo/test_solution.ipynb

Original Source

SSD-Tensorflow - https://github.com/balancap/SSD-Tensorflow