Match visitors to a photo collection
You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
Adrian Cochrane 54a9ab5dc1 Merge branch 'master' of https://github.com/alcinnz/Historical-Twin 7 months ago
assets Init. 7 months ago
Futura.ttc Init. 7 months ago
LICENSE Initial commit 7 months ago
README.md Init. 7 months ago
architecture.png Init. 7 months ago
closest-match.py Init. 7 months ago
describe_faces.py Init. 7 months ago
magic_camera.py Init. 7 months ago
magic_mirror.py Init. 7 months ago
recognition.py Init. 7 months ago
testpycairo.py Init. 7 months ago

README.md

Historical Twin

This software was written for the Tyree Photo Collection exhibit at the Nelson Provincial Museum, and I’m dumping the code online in case it’ll be helpful for anyone else.

To run it requires OpenFace installed, which I’ve successfully tested the installation instructions on elementary OS and Ubuntu.

Once OpenFace is installed run describe_faces.py over a directory of portrait photos you want Historical Twin to match visitors to. This may take a while, but once your done that directory would be configured to be loaded quickly into magic_camera.py or magic_mirror.py both of which alternative UIs for matching visitor faces to photos from the preconfigured collection.

You may also have to create a creditlines.tsv file which pairs filenames (excluding the 3 letter extension) in the first column to captions that should be displayed for the photo by magic_camera.py alongside the given photo.