After observing that some models were unfair, In this paper, we study the behaviour of the triplet loss and special efforts have been allocated to finding ways to mitigate show that it can be exploited to limit the biases created and those biases in the ML pipeline, with pre-processing perpetuated by machine learning models. Our fair classifier approaches [2, 23], which impact the data before it is uses the collapse of the triplet loss when its margin is ingested by the model, post-processing approaches [11], greater than the maximum distance between two points in which straighten out the model's output, and in-processing the latent space, in the case of stochastic triplet selection.

Siamese and triplet networks are useful to learn mappings from image to a compact Euclidean space where distances correspond to a measure of similarity [2]. Embeddings trained in such way can be used as features vectors for classification or few-shot learning tasks. Experiments were run in jupyter notebook. We'll go through learning supervised feature embeddings using different loss functions on MNIST dataset. This is just for visualization purposes, thus we'll be using 2-dimensional embeddings which isn't the best choice in practice.