LBM: Latent Bridge Matching for Fast Image-to-Image Translation

2025-03-10

AI Summary

To address low inference efficiency and poor multi-task generalization in image-to-image translation—such as object removal, normal/depth estimation, controllable relighting, and shadow generation—this paper proposes Latent Bridge Matching (LBM), the first adaptation of Bridge Matching to latent space for single-step, conditional diffusion-based image translation. Our key contributions are: (1) a latent-space bridge matching mechanism that aligns cross-domain distributions via optimal transport; (2) a lightweight conditional latent encoder-decoder architecture enabling fine-grained control; and (3) a unified framework supporting diverse translation tasks efficiently. Extensive experiments demonstrate state-of-the-art performance across multiple benchmarks, with single-step inference significantly faster than iterative diffusion or GAN-based methods. The code is publicly available.

Technology Category

• Computer Vision: Diffusion Models for Vision
• Machine Learning: Transfer, Domain Adaptation, Multi-Task Learning
• Natural Language Processing: Code Generation / Program Synthesis from Natural Language

Application Category

• Search and Retrieval-Augmented AI: Multilingual and cross-lingual Web search
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems

Abstract

In this paper, we introduce Latent Bridge Matching (LBM), a new, versatile and scalable method that relies on Bridge Matching in a latent space to achieve fast image-to-image translation. We show that the method can reach state-of-the-art results for various image-to-image tasks using only a single inference step. In addition to its efficiency, we also demonstrate the versatility of the method across different image translation tasks such as object removal, normal and depth estimation, and object relighting. We also derive a conditional framework of LBM and demonstrate its effectiveness by tackling the tasks of controllable image relighting and shadow generation. We provide an open-source implementation of the method at https://github.com/gojasper/LBM.

Problem

• Fast image-to-image translation using latent space
• Achieves state-of-the-art results in single inference step
• Versatile across tasks like object removal and relighting

Innovation

• Latent Bridge Matching for fast translation
• Single-step inference achieves top results
• Versatile across multiple image tasks