Sign inSign up

totti4ever/facebox-watchdog

By totti4ever

•Updated over 6 years ago

improved image of facebox with automation for teaching new faces plus saving and auto-loading state

Image
1

1.4K

totti4ever/facebox-watchdog repository overview

⁠Facebox with watchdog for training faces

⁠improved image of facebox:

  • automatically watching folder /training for new training images
  • automatically saves state-file and loads state-file at start-up
  • adds command train-all to container to retrain all images

⁠instructions

See the example docker-compose file for reference below, make sure you are binding the /training folder and keep the faces in a structure like this

  - training/
    - name1/
      - img-file1.jpg
      - image2.jpg
    - name2/
      - image1.jpg

⁠Example docker-compose.yml

version: "3"
services:
  facebox:
    image: totti4ever/facebox-watchdog
    container_name: facebox
    restart: unless-stopped
    hostname: facebox
    environment:
      - MB_KEY=[### ENTER YOUR API KEY HERE ###]
      - TZ=Europe/Berlin
    volumes:
      - ./training:/training

Tag summary

Content type

Image

Digest

Size

148.8 MB

Last updated

over 6 years ago

docker pull totti4ever/facebox-watchdog