Show HN: Finding dirty XTC with machine learning

I created a reverse image search for a harm reduction group as a side project. Many people don't test their MDMA, this at least makes it easier to search for past lab results. It's not a replacement for a chemical test.

It was my first project getting into machine learning.

It's made with 2 models, a Retinanet model for localization and a triplet loss model for generating image embeddings.

Those are then compared in an Annoy database and voila.

Feedback is appreciated. Thanks