Skip to content

Latest commit

 

History

History
61 lines (43 loc) · 1.13 KB

README.md

File metadata and controls

61 lines (43 loc) · 1.13 KB

WIP. First rust project created for learning the language.

equi - UI

dlib - for face detection and recognition

Instalation

read Dockerfile [TODO: fix dlib compilation in docker image]

Usage

Trainer

seed faces and data of few people to db:

cargo run train path_to_dir/

# where directory "path_to_dir" has structure like:
# path_to_dir/person_name/1.jpg
# path_to_dir/person_name/2.jpg
# path_to_dir/john_dou/1.jpg

Uses last dir as a person name.

CLI trainer

train app by using command line interface:

cargo run cli_trainer

Recognition

cargo run processor path/to/photo.jpg

recognizes all faces on given photo. Result is a vector of names and difference between closest matched face from database: [("John Dou", 0.25761012784916366)]

Recognition on video

cargo run video path/to/video.mp4

recognizes people from the first frame with detected faces.

UI

cargo run ui

Config

DATABASE_URL=./storage/db/database.sql
DATABASE_FOLDER=./storage/db/
MIGRATIONS_DIR=./migrations

can be set in .env file.

migration folder should include structure of db to create.