# MISO: Mutual Information Loss with Stochastic Style Representations for Multimodal Image-to-Image Translation

arXiv: https://arxiv.org/abs/1902.03938 · HTML: https://mlsanity.com/paper/1902.03938/

## Key facts

- Cited by 7 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 11 Feb 2019.

## Built on by (most cited first, 7 of 7)

- [Deepfake Style Transfer Mixture: a First Forensic Ballistics Study on Synthetic Images](https://mlsanity.com/paper/2203.09928/) (2022)
- [MFIM: Megapixel Facial Identity Manipulation](https://mlsanity.com/paper/2308.01536/) (2023)
- [VQ-Style: Disentangling Style and Content in Motion with Residual Quantized Representations](https://mlsanity.com/paper/2602.02334/) (2026)
- [Mutual-GAN: Towards Unsupervised Cross-Weather Adaptation with Mutual Information Constraint](https://mlsanity.com/paper/2106.16000/) (2021)
- [Separating Content and Style for Unsupervised Image-to-Image Translation](https://mlsanity.com/paper/2110.14404/) (2021)
- [SuperStyleNet: Deep Image Synthesis with Superpixel Based Style Encoder](https://mlsanity.com/paper/2112.09367/) (2021)
- [Enhanced Synthetic MRI Generation from CT Scans Using CycleGAN with Feature Extraction](https://mlsanity.com/paper/2310.20604/) (2023)

## Abstract (arXiv)

Unpaired multimodal image-to-image translation is a task of translating a given image in a source domain into diverse images in the target domain, overcoming the limitation of one-to-one mapping. Existing multimodal translation models are mainly based on the disentangled representations with an image reconstruction loss. We propose two approaches to improve multimodal translation quality. First, we use a content representation from the source domain conditioned on a style representation from the target domain. Second, rather than using a typical image reconstruction loss, we design MILO (Mutual Information LOss), a new stochastically-defined loss function based on information theory. This loss function directly reflects the interpretation of latent variables as a random variable. We show that our proposed model Mutual Information with StOchastic Style Representation(MISO) achieves state-of-the-art performance through extensive experiments on various real-world datasets.