Integrate CompareFace facial recognition

We have an existing Python based Windows desktop software where users can upload photos of lost people and found people. The software is for the purpose of searching for people who have been lost. The software is hosted in a server. And it's a multiuser software. It means many remote users have the software installed in their PC. Every user has an username and password that he/she uses to log in to his/her account in the software. The application also works as a website. So the users can alternatively log in to their account by visiting our server IP from a browser and then entering their username, password. So, our Windows software/website already has a database of photos of lost and found people. It uses MySQL database. The user account has 3 pages - "Matched photos", "Unmatched photos" and "All photos uploaded". When the user uploads a photo of a lost person or a found person, the photo is listed in "All photos uploaded" page. The "All photos uploaded" page in an user account only shows the photos uploaded by that specific user. We have an admin software too. Admin account can be accessed only from that software and not from the website. In the admin account, there are 3 pages too - "Matched photos", "Unmatched photos" and "All photos uploaded". The "All photos uploaded" page in the admin account shows all the photos of lost and found persons uploaded by all the users combined. The purpose of the software is to match photos of lost people against photos of found people, so that we can get a lost person back. When a photo of lost person is uploaded, it should be matched against all the photos of found persons saved in the database. And when a photo of a found person is uploaded, it should be matched against all the photos of lost persons saved in the database. If a match is found, the matched photos should be listed in "Matched photos" page of both users' accounts (the user who uploaded the photo of lost person and the user who uploaded the photo of found person), as well as the in "Matched photos" page of the admin panel. If a photo matches against multiple photos (very unlikely to happen), the set of matched photos should appear in the "Matched photos" page of all the users' accounts who have uploaded those photos and also in the "Matched photos" page of the admin panel. If a match is not found after uploading a photo, it should appear in the "Unmatched photos" page of the user's account, who have uploaded the photo as well as in the "Unmatched photos" page of the admin panel. However, as soon as an unmatched photo finds a match, it should be removed from the "Unmatched photos" page of the user's account, who have uploaded the photo as well as from the "Unmatched photos" page of the admin panel. So, you only have to create the "Matched photos" page and "Unmatched photos" page of the users accounts and admin panel. Rest of the functionalities are already built in our existing software. For matching photos, we need to use facial recognition Python libraries and models. InsightFace is an open-source Artificial Intelligence and Machine Learning based Python library that uses most recent and accurate facial recognition models such as ArcFace and RetinaFace. The accuracy of this solution is very high – 99.86% on the LFW dataset. [login to view URL] [login to view URL] To make your job easier, we recommend you to use CompreFace API. CompareFace is a free and open-source self-hosted REST API for InsightFace that can be started with one docker-compose command. The REST API allows you to easily integrate InsightFace into our system without prior machine learning skills. Additionally, it’s scalable. CompreFace has a simple UI for managing user roles and face collections. It gives a choice between the two most popular face recognition methods: FaceNet (LFW accuracy 99.65%) and InsightFace (LFW accuracy 99.86%). [login to view URL] [login to view URL] Software description attached as docx