kevintonb.

← All contributions

kevintonb

child-abuse-image-classifier

Image classifier to detect child abuse, trained on roughly 18,000 images. Won first place at the SRM-IET hackathon.

Image Classifier to detect child abuse , using Python and Tensorflow models trained with around 13000 training images.

This is the Google Drive link for Trained Data Models for

  1. Face Identification in the image
  2. Child Detection
  3. Abusive Content Detection
  4. The content , if found abusive is then blurred out in order to gain privacy

Also star ⭐️ the repo to gain access to the code ❤️

https://drive.google.com/drive/folders/1SnEcx2YtM-soUJPXsEqkLURcrX7mBZhi?usp=sharing

Place the above three downloaded folders in the folder where $run.sh is located

Installation: (Mac OS) brew install python (Python 2.7.3) pip install tensorflow (Tensorflow) pip install openCV (cv2)

To run: $ ./run.sh

This page mirrors the repository's own README, with GitHub's boilerplate and its contributor-facing sections removed. Read it at the source ↗