# An explanation method for Siamese neural networks

arXiv: https://arxiv.org/abs/1911.07702 · HTML: https://mlsanity.com/paper/1911.07702/

## Key facts

- Cited by 5 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 18 Nov 2019.

## Built on by (most cited first, 5 of 5)

- [Self-learn to Explain Siamese Networks Robustly](https://mlsanity.com/paper/2109.07371/) (2021)
- [DTWSSE: Data Augmentation with a Siamese Encoder for Time Series](https://mlsanity.com/paper/2108.09885/) (2021)
- [Identical Image Retrieval using Deep Learning](https://mlsanity.com/paper/2205.04883/) (2022)
- [Face Generation from Textual Features using Conditionally Trained Inputs to Generative Adversarial Networks](https://mlsanity.com/paper/2301.09123/) (2023)
- [A deep Natural Language Inference predictor without language-specific training data](https://mlsanity.com/paper/2309.02887/) (2023)

## Abstract (arXiv)

A new method for explaining the Siamese neural network is proposed. It uses the following main ideas. First, the explained feature vector is compared with the prototype of the corresponding class computed at the embedding level (the Siamese neural network output). The important features at this level are determined as features which are close to the same features of the prototype. Second, an autoencoder is trained in a special way in order to take into account the embedding level of the Si-amese network, and its decoder part is used for reconstructing input data with the corresponding changes. Numerical experiments with the well-known dataset MNIST illustrate the propose method.