Oneta: Multi-Style Image Enhancement Using Eigentransformation Functions

2025-06-30

AI Summary

This paper addresses multi-style image enhancement—including retouching, low-light enhancement, dehazing, underwater enhancement, white balance adjustment, and ISP modeling—by proposing Oneta, the first unified single-network framework. Methodologically, Oneta decouples enhancement into two stages: intensity enhancement and color correction, implemented via Y-Net and C-Net, respectively; these predict parameters of compact eigen-transform functions (eigenTFs) and color correction matrices. To support diverse enhancement styles, Oneta introduces learnable style tokens enabling end-to-end joint optimization over K distinct styles. Both stages employ efficient point-wise operators for lightweight, real-time processing. Evaluated across 30 benchmark datasets, Oneta simultaneously achieves state-of-the-art performance across all six enhancement tasks using a single model, demonstrating unprecedented generalization and practicality for multi-style image enhancement.

Technology Category

• Computer Vision: Multi-modal Vision
• Machine Learning: Multi-instance/Multi-view Learning
• Search and Optimization: Learning to Search

Application Category

• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Security and Privacy: Privacy-enhancing technologies
• Semantics and Knowledge: Representation, semantic annotation, enhancement, enrichments, access and/or integration of a variety of data on the Web

Abstract

The first algorithm, called Oneta, for a novel task of multi-style image enhancement is proposed in this work. Oneta uses two point operators sequentially: intensity enhancement with a transformation function (TF) and color correction with a color correction matrix (CCM). This two-step enhancement model, though simple, achieves a high performance upper bound. Also, we introduce eigentransformation function (eigenTF) to represent TF compactly. The Oneta network comprises Y-Net and C-Net to predict eigenTF and CCM parameters, respectively. To support $K$ styles, Oneta employs $K$ learnable tokens. During training, each style token is learned using image pairs from the corresponding dataset. In testing, Oneta selects one of the $K$ style tokens to enhance an image accordingly. Extensive experiments show that the single Oneta network can effectively undertake six enhancement tasks -- retouching, image signal processing, low-light image enhancement, dehazing, underwater image enhancement, and white balancing -- across 30 datasets.

Problem

• Proposes Oneta for multi-style image enhancement
• Uses eigentransformation functions for compact representation
• Handles six enhancement tasks across 30 datasets

Innovation

• Sequential intensity and color enhancement
• Compact eigentransformation function representation
• Multi-style tokens for diverse enhancement tasks