FLARE-BO: Fused Luminance and Adaptive Retinex Enhancement via Bayesian Optimisation for Low-Light Robotic Vision

2026-04-23

AI Summary

This work addresses the degradation of robotic visual perception under low-light conditions by proposing a training-free, adaptive image enhancement method. Built upon a Bayesian optimization framework, it uniquely integrates Retinex theory with a multidimensional enhancement strategy to jointly optimize eight parameters—gamma correction, illumination normalization, chromatic denoising, bilateral filtering, non-local means denoising, gray-world white balancing, and post-smoothing. To efficiently navigate the high-dimensional parameter space, the approach incorporates hypercube parameter normalization, Sobol quasi-random initialization, and a logarithmic expected improvement acquisition function. Experiments on the LOL dataset demonstrate that the proposed method significantly outperforms existing techniques that were not trained on this data, effectively enhancing low-light image quality and improving the robustness of robotic perception.

Technology Category

• Intelligent Robots: Learning & Optimization for ROB
• Computer Vision: Learning & Optimization for CV
• Machine Learning: Adversarial Learning & Robustness

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Responsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

Reliable visual perception under low illumination remains a core challenge for autonomous robotic systems, where degraded image quality directly compromises navigation, inspection, and various operations. A recent training free approach showed that Bayesian optimisation with Gaussian Processes can adaptively select brightness, contrast, and denoising parameters on a per-image basis, achieving competitive enhancement without any learned model. However, that framework is limited to three parameters, applies no illumination decomposition or white balance correction, and relies on Non-Local Means denoising, which tends to over smooth edges under noisy conditions. This paper proposes FLARE-BO (Fused Luminance and Adaptive Retinex Enhancement via Bayesian Optimisation), an extended framework that jointly optimises eight parameters spanning across gamma correction, LIME-style illumination normalisation, chrominance denoising, bilateral filtering, NLM denoising, Grey-World automatic white balance, and adaptive post smoothing. The search engine employs a unit hypercube parameter normalisation, objective standardisation, Sobol quasi-random initialisation, and Log Expected Improvement acquisition for principled exploration of the expanded space. Performance of the proposed method is benchmarked using the Low Light paired dataset (LOL) and results show marked improvements of the proposed method over existing methods that were not specifically trained using this dataset.

Problem

• low-light vision
• image enhancement
• robotic perception
• illumination decomposition
• white balance

Innovation

• Bayesian Optimisation
• Low-Light Enhancement
• Adaptive Retinex
• Illumination Normalisation
• Training-Free