Skip to content

An example notebook showing the use of a Siamese Neural Network with a triplet loss function trained on MNIST

License

Notifications You must be signed in to change notification settings

Trotts/Siamese-Neural-Network-MNIST-Triplet-Loss

Repository files navigation

Siamese Neural Network with Triplet Loss trained on MNIST

Cameron Trotter

This notebook builds an SNN to determine similarity scores between MNIST digits using a triplet loss function. The use of class prototypes at inference time is also explored.

This notebook is based heavily on the approach described in this Coursera course, which in turn is based on the FaceNet paper. Any uses of open-source code are linked throughout where utilised.

For an in-depth guide to understand this code, and the theory behind it, please see my article for Towards Data Science.

About

An example notebook showing the use of a Siamese Neural Network with a triplet loss function trained on MNIST

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published