adambielski/siamese-triplet
PyTorch library implementing siamese and triplet networks for learning compact embedding vectors from images.
Not currently ranked — collecting fresh signals.
star history
This repository provides implementations of siamese and triplet neural network architectures along with custom loss functions for contrastive learning. It includes online pair/triplet mining strategies that automatically select hard positive and negative examples within mini-batches to improve embedding quality. The code provides dataset wrappers for MNIST, batch samplers for balanced training, and a unified fit function for training these embedding networks.
Frequently asked
- What is adambielski/siamese-triplet?
- PyTorch library implementing siamese and triplet networks for learning compact embedding vectors from images.
- Is siamese-triplet open source?
- Yes — adambielski/siamese-triplet is open source, released under the BSD-3-Clause license.
- What language is siamese-triplet written in?
- adambielski/siamese-triplet is primarily written in Python.
- How popular is siamese-triplet?
- adambielski/siamese-triplet has 3.2k stars on GitHub.
- Where can I find siamese-triplet?
- adambielski/siamese-triplet is on GitHub at https://github.com/adambielski/siamese-triplet.