# Revisiting Adversarial Autoencoder for Unsupervised Word Translation with Cycle Consistency and Improved Training

## TL;DR

This paper improves adversarial autoencoder methods for unsupervised word translation by introducing cycle consistency and input reconstruction regularizations, resulting in more stable training and superior performance across diverse language pairs.

## Contribution

It proposes two novel extensions to adversarial autoencoders, enhancing stability and translation accuracy in unsupervised bilingual dictionary learning.

## Key findings

- More robust training process.
- Outperforms recent adversarial and non-adversarial methods.
- Effective across European, non-European, and low-resource languages.

## Abstract

Adversarial training has shown impressive success in learning bilingual dictionary without any parallel data by mapping monolingual embeddings to a shared space. However, recent work has shown superior performance for non-adversarial methods in more challenging language pairs. In this work, we revisit adversarial autoencoder for unsupervised word translation and propose two novel extensions to it that yield more stable training and improved results. Our method includes regularization terms to enforce cycle consistency and input reconstruction, and puts the target encoders as an adversary against the corresponding discriminator. Extensive experimentations with European, non-European and low-resource languages show that our method is more robust and achieves better performance than recently proposed adversarial and non-adversarial approaches.