FaceMatch is a system for identifying facial matches within an image database. With FaceMatch, users can create a database of images of people and, by uploading a new image, quickly find any matches for the person of interest within the database.
Built with a client-server architecture using Flask-ML, FaceMatch provides structured support for efficient client-server communication tailored to ML applications.
- Python 3.12
- Virtual environment support (recommended but optional)
git clone https://github.com/RigvedManoj/FaceMatch.git
cd FaceMatch
python -m venv facematch-env
source facematch-env/bin/activate
# On Windows: facematch-env\Scripts\activate
conda create -n facematch-env python=3.12
conda activate facematch-env
Run below command from root directory of project.
pip install -r requirements.txt
Sample dataset to test the model: The images in the \resources\sample_images folder can be used as the database, and \resources\test_image.jpg can be used as the query image to test the model.
Run all below commands from root directory of project.
python -m src.facematch.face_match_server
python -m src.Sample_Client.sample_bulk_upload_client --directory_paths <path_to_directory_of_images> --database_name <database_name>
Note: The name of the database could be a new database you wish to create or an existing database you wish to upload to.
Run with Sample images directory: (Requires absolute path of directory)
python -m src.Sample_Client.sample_bulk_upload_client --directory_paths <path_to_project>\resources\sample_images --database_name test_database
python -m src.Sample_Client.sample_find_face_client --file_paths <path_to_image> --database_name <database_name> --similarity_threshold <similarity_threshold>
Note: The name of the database needs to be an existing database you wish to query. The default similarity threshold, 0.45 is used if no similarity threshold is provided.
Run with Sample test image: (Requires absolute path of image)
python -m src.Sample_Client.sample_find_face_client --file_paths <path_to_project>\resources\test_image.jpg --database_name test_database --similarity_threshold 0.5
Run below command from root directory of project.
python -m src.facematch.face_match_server
-
Install Rescue-Box from link
-
Open Rescue-Box-Desktop and resgiter the model by adding the server IP address and port number in which the server is running.
-
Choose the model from list of available models under the MODELS tab.
-
Checkout the Inspect page to learn more about using the model.
We conduct a series of experiments to evaluate the performance and accuracy of FaceMatch. The experiments are conducted on a dataset of 1,680 images of people from the Labelled Faces in the Wild (LFW) dataset.
We evaluate FaceMatch using the following metrics:
In this type of testing, all test images have a corresponding match in the database. If match is returned, it is True, otherwise False.
- Upload Time: The time taken to upload images to the database.
- Search Time per Image: The time taken to find matches for a single image.
- Faiss Accuracy (Top-n): The percentage of images in the database that have at least one match within the top-n results using FAISS search algorithm. Atleast one match in top-n is considered a positive match, meaning that the person of interest is identified within the image.
| Face Recognition Model | Number of Faces in Database | Upload Time (in seconds) | Search Time per Image (in seconds) | Top-n | Faiss Accuracy |
|---|---|---|---|---|---|
| ArcFace | 1680 | 242 | < 1.5 | 10 | 93.2 |
| FaceNet | 1680 | 463 | < 2 | 10 | 94 |
| VGGFace | 1680 | 650 | < 3 | 10 | 90 |
In this type of testing, half the test images have a corresponding match in the database, the other half does not. This helps measure True Positive, False Positive, True Negative and False Negative.
| Face Recognition Model | Number of Faces in Database | Threshold | Accuracy | Precision | Recall |
|---|---|---|---|---|---|
| ArcFace | 840 | 0.48 | 77 | 79 | 73 |
| FaceNet | 840 | 0.63 | 75 | 75 | 75 |
The above metrics were calculated using the following system configuration:
- OS: Windows 11 Pro 64-bit
- Processor: AMD Ryzen 7 7735HS with Radeon Graphics (16 CPUs), ~3.2 GHz
- RAM: 32 GB
Check out Testing README for complete details on dataset and testing.
Check out this document for more details on the face detection and recognition models, datasets and testing.
Run all below commands from root directory of project.
python -m unittest discover test
python -m unittest test.<test_file_name>