IM-LUT: Interpolation Mixing Look-Up Tables for Image Super-Resolution

2025-07-14

AI Summary

Existing arbitrary-scale image super-resolution (ASISR) methods face a fundamental trade-off: lookup-table (LUT)-based approaches support only fixed scaling factors, whereas implicit neural representations incur prohibitive computational and memory overhead. To address this, we propose IM-LUT—a highly efficient, lightweight ASISR framework. First, we design a learnable Interpolation-Mixing Network (IM-Net) that dynamically fuses multiple interpolation kernels based on local image features and the target scale. Second, we distill IM-Net offline into a compact, model-free lookup table (IM-LUT), enabling real-time, CPU-only inference via pure table lookups. To our knowledge, IM-LUT is the first method to integrate adaptive interpolation mixing with LUT-based deployment. Extensive experiments demonstrate that IM-LUT consistently outperforms state-of-the-art ASISR methods across multiple benchmarks, achieving superior quality-efficiency trade-offs and unprecedented practicality for edge devices.

Technology Category

• Computer Vision: Representation Learning for Vision
• Machine Learning: Large Multimodal Models (LMMs)
• Search and Optimization: Learning to Search

Application Category

• Search and Retrieval-Augmented AI: Efficiency and scalability of Web search engines
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Super-resolution (SR) has been a pivotal task in image processing, aimed at enhancing image resolution across various applications. Recently, look-up table (LUT)-based approaches have attracted interest due to their efficiency and performance. However, these methods are typically designed for fixed scale factors, making them unsuitable for arbitrary-scale image SR (ASISR). Existing ASISR techniques often employ implicit neural representations, which come with considerable computational cost and memory demands. To address these limitations, we propose Interpolation Mixing LUT (IM-LUT), a novel framework that operates ASISR by learning to blend multiple interpolation functions to maximize their representational capacity. Specifically, we introduce IM-Net, a network trained to predict mixing weights for interpolation functions based on local image patterns and the target scale factor. To enhance efficiency of interpolation-based methods, IM-Net is transformed into IM-LUT, where LUTs are employed to replace computationally expensive operations, enabling lightweight and fast inference on CPUs while preserving reconstruction quality. Experimental results on several benchmark datasets demonstrate that IM-LUT consistently achieves a superior balance between image quality and efficiency compared to existing methods, highlighting its potential as a promising solution for resource-constrained applications.

Problem

• Enables arbitrary-scale image super-resolution efficiently
• Reduces computational cost of neural representations
• Balances image quality and processing speed

Innovation

• Blends multiple interpolation functions for ASISR
• Uses LUTs to replace expensive operations
• Lightweight fast inference on CPUs