Faiss + Numba + SciPy + FastAPI + LMDB
Uses DCT-based phash (576 bit)
Supported operations: add new image, delete image, find similar images by image file, find similar images by image id
pip3 install -r requirements.txt
generate_phashes.py ./path_to_img_folder -> generates features
--use_int_filenames_as_id=0 - images get sequential ids
--use_int_filenames_as_id=1 - image id is parsed from filename ("123.jpg" -> 123)
add_to_index.py -> adds features from lmdb to Flat index
phash_web.py -> web microservice
GET_FILENAMES=1 phash_web.py -> when searching, include filename in search results
DOCKER:
build image - docker build -t qwertyforce/phash_web:1.0.0 --network host -t qwertyforce/phash_web:latest ./
run interactively - docker run -ti --rm -p 127.0.0.1:33336:33336 --network=ambience_net --mount type=bind,source="$(pwd)"/data,target=/app/data --name phash_web qwertyforce/phash_web:1.0.0
run as deamon - docker run -d --rm -p 127.0.0.1:33336:33336 --network=ambience_net --mount type=bind,source="$(pwd)"/data,target=/app/data --name phash_web qwertyforce/phash_web:1.0.0