WWE-UIE: A Wavelet & White Balance Efficient Network for Underwater Image Enhancement

Underwater Image Enhancement (UIE) aims to restore visibility and correct color distortions caused by light absorption and scattering in aquatic environments. Existing deep learning approaches often suffer from high computational overhead or limited generalization due to the lack of explicit domain priors. In this work, we propose WWE-UIE, a compact and efficient enhancement network that integrates three interpretable priors: white balance correction to mitigate color attenuation, a wavelet-based enhancement block (WEB) for multi-scale frequency decomposition, and a gradient-aware module (SGFB) to preserve edge sharpness. We further incorporate the HVI color space to decouple chromatic and intensity information, enhancing color fidelity in challenging underwater scenes. Extensive experiments on benchmark datasets demonstrate that WWE-UIE achieves competitive performance with significantly fewer parameters and FLOPs, enabling real-time inference on resource-limited platforms. Ablation studies and visualizations confirm the contribution of each component. Moreover, our prior-guided framework generalizes well to other degradation domains, such as low-light and foggy scenes, highlighting its versatility for practical and time-sensitive image restoration applications.