Learning Pixel-adaptive Multi-layer Perceptrons for Real-time Image Enhancement

2025-07-16

AI Summary

Existing bilateral grid methods are constrained by linear transformations, limiting their ability to model complex color relationships; meanwhile, global shared-parameter MLPs lack spatial adaptability. To address these limitations, we propose a pixel-adaptive bilateral grid MLP framework: MLP weights are embedded into the bilateral grid, enabling each pixel to dynamically generate its own nonlinear mapping function. We further introduce a multi-channel guidance map to decompose the parameter sub-grids, achieving joint spatial-intensity modeling and channel-specific parameter decoupling. The proposed method preserves full-resolution real-time inference capability while significantly improving color enhancement fidelity and detail restoration quality across multiple public benchmarks. To our knowledge, this is the first approach to realize pixel-level nonlinear adaptive mapping on bilateral grids.

Technology Category

• Machine Learning: Multimodal Learning
• Computer Vision: Multi-modal Vision
• Natural Language Processing: Language Grounding & Multi-modal NLP

Application Category

• User Modeling, Personalization and Recommendation: User modeling for targeted and personalized online advertising
• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Deep learning-based bilateral grid processing has emerged as a promising solution for image enhancement, inherently encoding spatial and intensity information while enabling efficient full-resolution processing through slicing operations. However, existing approaches are limited to linear affine transformations, hindering their ability to model complex color relationships. Meanwhile, while multi-layer perceptrons (MLPs) excel at non-linear mappings, traditional MLP-based methods employ globally shared parameters, which is hard to deal with localized variations. To overcome these dual challenges, we propose a Bilateral Grid-based Pixel-Adaptive Multi-layer Perceptron (BPAM) framework. Our approach synergizes the spatial modeling of bilateral grids with the non-linear capabilities of MLPs. Specifically, we generate bilateral grids containing MLP parameters, where each pixel dynamically retrieves its unique transformation parameters and obtain a distinct MLP for color mapping based on spatial coordinates and intensity values. In addition, we propose a novel grid decomposition strategy that categorizes MLP parameters into distinct types stored in separate subgrids. Multi-channel guidance maps are used to extract category-specific parameters from corresponding subgrids, ensuring effective utilization of color information during slicing while guiding precise parameter generation. Extensive experiments on public datasets demonstrate that our method outperforms state-of-the-art methods in performance while maintaining real-time processing capabilities.

Problem

• Overcoming linear affine limitations in image enhancement
• Addressing localized variations in MLP-based methods
• Combining bilateral grids and MLPs for real-time processing

Innovation

• Pixel-adaptive MLPs for dynamic color mapping
• Bilateral grid with decomposed MLP subgrids
• Multi-channel guidance for precise parameter generation