adult-image-detector

by open-dating

GitHub Readme.md

adult-image-detector

Deploy Docker Cloud Build Status

Use deep neural networks and other algos for detect nude images.

Try detection

Usage

Exec:

curl -i -X POST -F "image=@Daddy_Lets_Me_Ride_His_Cock_preview_720p.mp4.jpg" http://localhost:9191/api/v1/detect

Result:

{
  "app_version":"0.2.0",
  "open_nsfw_score":0.81577397,
  "an_algorithm_for_nudity_detection": true,
  "image_name":"Daddy_Lets_Me_Ride_His_Cock_preview_720p.mp4.jpg"
}

Docker

Run

docker run -p 9191:9191 opendating/adult-image-detector

Build

git clone https://github.com/open-dating/adult-image-detector --recursive
docker build -t adult-image-detector .

Development

git clone https://github.com/open-dating/adult-image-detector --recursive
cd docker/dev
docker-compose up

Test

cd docker/test
docker-compose up

Install to heroku

Use deploy button or:

fork, create app and change stack to container

heroku stack:set container

Requirements

Go 1.11

opencv 3.4.1

Development without docker

Recursive clone that repo:

git clone https://github.com/open-dating/adult-image-detector --recursive

or manually install submodules:

git submodule init
git submodule update

Install dependencies:

# get package manager
go get -u github.com/kardianos/govendor

# install dependencies
govendor sync

Install opencv 3.4.1

Run hot reload with fresh:

go get github.com/pilu/fresh
fresh