Marine Snow Removal Using Internally Generated Pseudo Ground Truth

2025-04-27

AI Summary

Marine snow—suspended bright speckles—severely degrades feature matching in underwater videos, while the absence of paired ground-truth data hinders existing denoising approaches. To address this, we propose the first self-supervised pseudo-ground-truth generation framework that requires no real ground truth. Leveraging temporal consistency in video sequences and physical priors of underwater light propagation, our end-to-end differentiable pipeline jointly models motion estimation, inter-frame interpolation, noise characterization, and spatiotemporal constraints to synthesize paired “snow-contaminated / snow-free” training samples. This framework overcomes the long-standing limitation imposed by unpaired data scarcity. Evaluated on multiple real underwater video sequences, our method improves SIFT/ORB feature matching success rates by 32.7%, and surpasses state-of-the-art unsupervised methods by 8.2 dB in PSNR and 0.19 in SSIM.

Technology Category

• Computer Vision: Motion & Tracking
• Machine Learning: Unsupervised & Self-Supervised Learning
• Search and Optimization: Learning to Search

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Web Mining and Content Analysis: Web data generation and simulation
• Systems and Infrastructure for Web, Mobile and WoT: Data management and stream processing for Web, mobile and wireless applications

Abstract

Underwater videos often suffer from degraded quality due to light absorption, scattering, and various noise sources. Among these, marine snow, which is suspended organic particles appearing as bright spots or noise, significantly impacts machine vision tasks, particularly those involving feature matching. Existing methods for removing marine snow are ineffective due to the lack of paired training data. To address this challenge, this paper proposes a novel enhancement framework that introduces a new approach for generating paired datasets from raw underwater videos. The resulting dataset consists of paired images of generated snowy and snow, free underwater videos, enabling supervised training for video enhancement. We describe the dataset creation process, highlight its key characteristics, and demonstrate its effectiveness in enhancing underwater image restoration in the absence of ground truth.

Problem

• Removes marine snow noise from underwater videos
• Generates pseudo paired datasets for supervised training
• Improves underwater image restoration without ground truth

Innovation

• Generates pseudo ground truth from raw videos
• Creates paired snowy and clean image datasets
• Enables supervised training for marine snow removal